# Overview Source: https://docs.oneperfectslice.ai/accounts/overview OnePerfectSlice is structured around organizations and teams. Organizations own integrations and billing. Teams scope calls, summaries, scorecards, and slice runs — so each team only sees its own data. ## Organizations and teams * An **organization** is your company's account. It owns integrations (CRM, Notion), billing, and API keys. * A **team** is a group within your organization. Each team has its own calls, summaries, scorecards, and slice runs. Most data is **team-scoped** — a team only sees its own data. CRM data (deals, companies, contacts) is **org-scoped** because the CRM connection is shared. ## How it's structured ``` Organization (your company) ├── Team A (e.g., Enterprise Sales) │ ├── Members │ ├── Calls, summaries, scorecards │ └── Slice runs and evidence ├── Team B (e.g., Mid-Market Sales) │ ├── Members │ ├── Calls, summaries, scorecards │ └── Slice runs and evidence └── Shared across org ├── CRM data (HubSpot deals, companies, contacts) ├── Integrations (Gong, Fireflies, Fathom, Notion) ├── Billing └── API keys (each scoped to one team) ``` ## Users and membership * A user can belong to **multiple teams** within the same organization * A user can belong to **multiple organizations** * Each user has a **primary team** per organization — this is their default context when they log in * Users are added via invite (email) or can be added directly by an Org Admin ## API keys and scoping API keys are created at the org level but scoped to a single team. A key for Team A can only access Team A's data. See [Authentication](/api/authentication) for details. # Roles & Permissions Source: https://docs.oneperfectslice.ai/accounts/roles-and-permissions OnePerfectSlice uses role-based access control at two levels — organization and team. Higher roles inherit all permissions from the roles below them. ## Roles | Role | Scope | Description | | -------------- | ------------ | ---------------------------------------------------------------------------------------- | | **Org Admin** | Organization | Full access — manage users, settings, integrations, billing, and API keys across the org | | **Team Admin** | Team | Manage team members and team-level settings | | **Member** | Team | Standard access — view and interact with calls, summaries, scorecards, and slices | | **Viewer** | Team | Read-only access | | **Billing** | Organization | View and manage billing only | ## What each role can do | Action | Org Admin | Team Admin | Member | Viewer | | ------------------------------------------ | :-------: | :--------: | :----: | :----: | | View calls, summaries, scorecards | Yes | Yes | Yes | Yes | | Run slices | Yes | Yes | Yes | No | | Upload calls | Yes | Yes | Yes | No | | Delete calls | Yes | Yes | No | No | | Manage team members | Yes | Yes | No | No | | Connect integrations (Gong, HubSpot, etc.) | Yes | No | No | No | | Create API keys | Yes | No | No | No | | Manage billing | Yes | No | No | No | | Manage org settings | Yes | No | No | No | | Access all teams in org | Yes | No | No | No | ## How roles are assigned * When you create an organization, you become the **Org Admin** * When you invite someone, you choose their role * **Org-level roles** (Org Admin) grant access across all teams in the organization * **Team-level roles** (Team Admin, Member, Viewer) are scoped to a specific team * A user can have different roles on different teams # List CRM custom fields Source: https://docs.oneperfectslice.ai/api-reference/filters/list-crm-custom-fields /api/openapi.yaml get /api/public/v1/filters/crm-fields Returns the CRM fields configured for filtering in your organization, including field type, label, and valid options for enumeration fields. Use this to build `crmCustomFieldFilters` in your requests. # List CRM deal stages Source: https://docs.oneperfectslice.ai/api-reference/filters/list-crm-deal-stages /api/openapi.yaml get /api/public/v1/filters/crm-deal-stages Returns all CRM deal stages for your organization with call counts. Use the results to populate `crmDealStages` filters in slice runs and post searches. # Search call owners Source: https://docs.oneperfectslice.ai/api-reference/filters/search-call-owners /api/openapi.yaml get /api/public/v1/filters/call-owners Search for call owners on your team by name. Returns matching owners with their email and call count. Use the results to populate `callOwnerEmails` filters. **Related:** [Search posts](/api-reference/posts/search-posts) · [Run a slice](/api-reference/slices/run-a-slice) Pass `q` for typeahead search, or omit it to get all call owners. Use the returned `email` values in `callOwnerEmails` filters on slice runs and post searches. # Search company domains Source: https://docs.oneperfectslice.ai/api-reference/filters/search-company-domains /api/openapi.yaml get /api/public/v1/filters/company-domains Search for company domains associated with calls on your team. Returns matching domains with call counts. Use the results to populate `companyDomains` filters. # Search CRM companies Source: https://docs.oneperfectslice.ai/api-reference/filters/search-crm-companies /api/openapi.yaml get /api/public/v1/filters/crm-companies Search CRM companies by name across your organization. Returns matching companies with call counts. CRM data is shared across all teams in your org. # Search CRM deals Source: https://docs.oneperfectslice.ai/api-reference/filters/search-crm-deals /api/openapi.yaml get /api/public/v1/filters/crm-deals Search CRM deals by name across your organization. Returns matching deals with stage, amount, and call count. CRM data is shared across all teams in your org. # Get a post Source: https://docs.oneperfectslice.ai/api-reference/posts/get-a-post /api/openapi.yaml get /api/public/v1/posts/{id} Returns the full detail for a single post, including the structured result payload, call metadata, owner, and team information. Use `fileId` (not `postId`) to link summaries and scorecards for the same call — a single call can produce both a summary post and a scorecard post. # List post filters Source: https://docs.oneperfectslice.ai/api-reference/posts/list-post-filters /api/openapi.yaml get /api/public/v1/posts/filters Returns the available filter keys and their valid values for post search. Use this to discover what filters your team can apply before calling POST /posts/search. Call this first when building a filter UI. The response tells you exactly which filter keys exist and what values are valid for each. # Search posts Source: https://docs.oneperfectslice.ai/api-reference/posts/search-posts /api/openapi.yaml post /api/public/v1/posts/search Search AI-generated summaries and scorecards for your team. Filter by date, call type, rep, company, CRM deal, and more. Send an empty body `{}` to return recent posts. Set `detail: true` to get full post payloads, or leave it `false` (default) for compact summaries with snippets. **Related:** [Get a post](/api-reference/posts/get-a-post) · [List post filters](/api-reference/posts/list-post-filters) Start with `detail: false` (the default) for faster, smaller responses. Only set `detail: true` when you need the full structured payload for a specific post. # List call types Source: https://docs.oneperfectslice.ai/api-reference/reference-data/list-call-types /api/openapi.yaml get /api/public/v1/call-types Returns all active call types for your team (e.g., "Discovery Call", "Demo", "QBR"), including a count of how many calls have each type. Use the `key` values from this response to populate `callTypes` filters in slice runs and evidence searches. # List elements Source: https://docs.oneperfectslice.ai/api-reference/reference-data/list-elements /api/openapi.yaml get /api/public/v1/elements Returns all active analysis elements for your team (e.g., "Budget Confirmation", "Competitive Mention"). Elements are the specific things a slice evaluates in your calls. Use the `key` values from this response as `elementKeys` when searching evidence across runs. # Get run details Source: https://docs.oneperfectslice.ai/api-reference/slice-runs/get-run-details /api/openapi.yaml get /api/public/v1/slice-runs/{id} Returns the current status, progress, and results for a slice run. Use this to poll an in-progress run — check `suggestedPollAfterSeconds` for the recommended wait time. When `status` is `completed`, the `result` field contains structured output conforming to the slice's output schema. **Related:** [Get run evidence](/api-reference/slice-runs/get-run-evidence) · [Search evidence](/api-reference/slice-runs/search-evidence) Watch the `phase` field for granular progress: `queued` → `gathering_evidence` → `analyzing_calls` → `synthesizing` → `completed`. Show this to users if building a progress UI. # Get run evidence Source: https://docs.oneperfectslice.ai/api-reference/slice-runs/get-run-evidence /api/openapi.yaml get /api/public/v1/slice-runs/{id}/evidence Returns evidence items for a completed run, grouped by element. Each item includes a direct quote, source reference, and confidence score. Capped at 20 items per element — use the `count` field to see the true total. This endpoint only works for completed runs. If the run is still in progress, you'll get a `409 RUN_NOT_COMPLETED` error. Poll [Get run details](/api-reference/slice-runs/get-run-details) first to confirm the run has finished. # List runs Source: https://docs.oneperfectslice.ai/api-reference/slice-runs/list-runs /api/openapi.yaml get /api/public/v1/slice-runs Returns a paginated list of slice runs for your team, ordered by most recent first. Filter by slice key or status (pending, running, completed, failed). Use `status=running` to check how many runs are active before starting a new one — there's a limit of 5 concurrent runs per team. # Search evidence Source: https://docs.oneperfectslice.ai/api-reference/slice-runs/search-evidence /api/openapi.yaml post /api/public/v1/evidence/search Search for evidence across multiple completed runs. Filter by element keys, text query, call types, call owners, and date range. Useful for finding specific quotes or patterns across analyses. Unlike [Get run evidence](/api-reference/slice-runs/get-run-evidence) which returns evidence for a single run, this endpoint searches across all your completed runs — useful for trend analysis or finding recurring themes. # Get slice details Source: https://docs.oneperfectslice.ai/api-reference/slices/get-slice-details /api/openapi.yaml get /api/public/v1/slices/{key} Returns the complete definition for a single slice, including its output schema, preferred model, and evidence grouping strategy. Use this to understand what a slice produces before running it. The `outputSchema` field tells you the exact shape of the structured result you'll get when a run completes. Use it to plan how your integration handles the output. # List slices Source: https://docs.oneperfectslice.ai/api-reference/slices/list-slices /api/openapi.yaml get /api/public/v1/slices Returns all AI analysis templates available to your team, including each slice's key, title, business objective, and the elements it evaluates. **Related:** [Get slice details](/api-reference/slices/get-slice-details) · [Preview matched call count](/api-reference/slices/preview-matched-call-count) This is usually your first API call. Browse the available slices, then use [Get slice details](/api-reference/slices/get-slice-details) to see the full output schema before running one. # Preview matched call count Source: https://docs.oneperfectslice.ai/api-reference/slices/preview-matched-call-count /api/openapi.yaml post /api/public/v1/slices/{key}/preview-count Returns how many calls match your filters without starting an analysis run. Use this to check scope and cost before committing to a run. Dates are required; all other filters are optional. Always preview before running. This call is free — it only counts matching calls without using any AI credits. # Run a slice Source: https://docs.oneperfectslice.ai/api-reference/slices/run-a-slice /api/openapi.yaml post /api/public/v1/slices/{key}/run Starts an asynchronous AI analysis run for the given slice. Returns `202 Accepted` with a `runId` you can use to poll for results via GET /slice-runs/{id}. Pass an `Idempotency-Key` header to safely retry — same key with same arguments returns the original run, same key with different arguments returns `409`. **Related:** [Get run details](/api-reference/slice-runs/get-run-details) · [Get run evidence](/api-reference/slice-runs/get-run-evidence) · [List runs](/api-reference/slice-runs/list-runs) Runs are asynchronous. After starting a run, poll [Get run details](/api-reference/slice-runs/get-run-details) using the returned `runId`. Check `suggestedPollAfterSeconds` for the recommended wait time between polls. # Authentication Source: https://docs.oneperfectslice.ai/api/authentication Authenticate with the OnePerfectSlice API using team-scoped Bearer token API keys. Learn how auth works, manage keys, handle errors, and follow security best practices. Every request to the OnePerfectSlice API requires a Bearer token in the `Authorization` header. ```bash theme={null} Authorization: Bearer sk_your_api_key ``` API keys start with `sk_` and are created in Org Settings. Don't have one yet? Follow the [Getting Started](/api/introduction) guide. ## How authentication works Each API key is scoped to a **single team**. When you make a request, OnePerfectSlice verifies the key and returns only data visible to that team. * **One key, one team** — a key created for Team A cannot read Team B's data * **Org-level isolation** — requests for resources in a different organization return `404` (not `403`) to prevent leaking whether a resource exists ## Managing your API keys Org Admins can manage keys from **Org Settings → API Keys**: | Action | What it does | | ---------- | --------------------------------------------------------------------------------------- | | **Create** | Generate a new key for a specific team. The key is shown once — copy it immediately. | | **View** | See all keys across your org, including name, team, and last-used timestamp | | **Revoke** | Permanently disable a key. Takes effect immediately — any in-flight requests will fail. | OnePerfectSlice never stores your raw key. If you lose a key, create a new one and revoke the old one. ## Authentication errors If authentication fails, the API returns a `401` with an error code like `MISSING_TOKEN`, `INVALID_TOKEN`, `TOKEN_REVOKED`, or `TOKEN_EXPIRED`. See the [Errors page](/api/errors#authentication-401) for the full reference. ## Security best practices * **Use environment variables** — store keys in `.env` or your secrets manager, never in source code * **Create separate keys** for each use case (production, staging, MCP server, Zapier) * **Rotate keys regularly** — create a new key, update your integrations, then revoke the old one * **Revoke compromised keys immediately** — revocation takes effect on the next request * **Never expose keys** in client-side code, git repos, logs, or error messages # Data Models Source: https://docs.oneperfectslice.ai/api/data-models Core entities and how they relate to each other. ## Entity relationships ``` Slice (analysis template) └── Run (one execution of a slice) ├── Result (structured output conforming to slice's outputSchema) └── Evidence (quotes from transcripts and CRM records) Call Type (category of call, e.g. "Discovery Call") └── Post (AI-generated output from a processed call) ├── Summary (high-level call recap) └── Scorecard (element-by-element performance evaluation) ``` ## Slices A slice is a pre-built AI analysis template. Each slice has a business objective (pipeline, deals, expand, or product) and evaluates conversations against a set of **elements** (e.g., "budget confirmation", "competitive mention"). | Field | Description | | ------------------ | ----------------------------------------------------------- | | `key` | Unique identifier (e.g., `high_intent_deals`) | | `title` | Human-readable name | | `objective` | Business category: `pipeline`, `deals`, `expand`, `product` | | `elementKeys` | Elements this slice evaluates | | `outputSchema` | JSON schema defining the structured output | | `preferredModel` | LLM model used for analysis | | `evidenceGrouping` | How evidence is organized: `ByElement` or `ByAccount` | ## Runs A run is one execution of a slice. Runs are asynchronous — you start one and poll for completion. | Field | Description | | --------------------------- | --------------------------------------------------------------------------------------------------- | | `id` | Unique run identifier | | `status` | `pending`, `running`, `completed`, `failed` | | `phase` | Execution phase: `queued` → `gathering_evidence` → `analyzing_calls` → `synthesizing` → `completed` | | `matchedCallCount` | Number of calls matching the filters | | `callsAnalyzed` | Number of calls analyzed so far | | `result` | Structured output (null until completed) | | `suggestedPollAfterSeconds` | Recommended polling interval | ## Evidence Evidence items are the supporting data behind a run's results. Each item is a quote from a transcript or CRM record, grouped by element. | Field | Description | | -------------- | --------------------------------------------------------- | | `quote` | Extracted text from the source | | `source.type` | Source type (e.g., `transcript`, `crm_deal`) | | `source.id` | Source identifier (e.g., `file:98123`) | | `source.label` | Human-readable label (e.g., "Acme Corp — Discovery Call") | | `confidence` | Confidence score (0.0 to 1.0) | Evidence is capped at **20 items per element** per run. ## Posts Posts are AI-generated outputs from processed calls. Two types: | Type | Description | | ------------- | ------------------------------------------------------------------ | | **Summary** | High-level call recap with key topics, decisions, and next steps | | **Scorecard** | Element-by-element performance evaluation with scores and coaching | | Field | Description | | ---------------- | --------------------------------------------------------------------------------- | | `postId` | Post identifier | | `fileId` | Underlying call identifier (use to deduplicate summary + scorecard for same call) | | `templateType` | `summary` or `scorecard` | | `title` | Call title | | `result` | Structured post payload | | `callOwnerEmail` | Who ran the call | | `companyDomain` | External participant's company | ## Call Types Call types categorize your conversations (e.g., "Discovery Call", "Demo", "QBR"). | Field | Description | | ------------- | ------------------------------------------ | | `key` | Unique identifier (e.g., `discovery_call`) | | `label` | Human-readable name | | `description` | What this call type represents | | `numCalls` | Number of calls with this type | ## Elements Elements are the specific things a slice evaluates (e.g., "budget confirmation", "competitive mention"). | Field | Description | | ------------- | ----------------------------------------------- | | `key` | Unique identifier (e.g., `budget_confirmation`) | | `label` | Human-readable name | | `description` | What this element detects | ## Scoping | Entity | Scope | Meaning | | ----------------------------------------- | -------- | -------------------------------------- | | Slices, Runs, Posts, Call Types, Elements | **Team** | Visible to the API key's team only | | CRM Deals, Companies, Stages, Fields | **Org** | Shared across teams (CRM is org-level) | # Endpoints Source: https://docs.oneperfectslice.ai/api/endpoints The OnePerfectSlice API has 20 endpoints across five groups — slices, runs, posts, filters, and reference data. Here's how they're organized, the typical integration workflow, and what each endpoint does. Most integrations follow the same pattern: discover what analyses are available, run one, and retrieve the results. The 20 endpoints are organized into five groups that map to this workflow. ## Typical workflow ``` List slices → see what analyses your team has access to Preview count → check how many calls match your filters (free, no AI cost) Run a slice → kick off async AI analysis (returns 202 immediately) Poll the run → check status until completed Get results → pull structured output and supporting evidence ``` This is the most common path, but not the only one. You can search posts directly if you just need summaries, browse past runs without starting new ones, or use the filter endpoints to build a selection UI. ## Slices Slices are AI analysis templates. Each one targets a specific business question — like "which deals show buying intent?" or "where are reps missing discovery questions?" — and evaluates your calls against a set of scoring elements. | Method | Path | What it does | | ------ | ----------------------------- | -------------------------------------------------------------------- | | GET | `/slices` | List every slice your team has access to | | GET | `/slices/{key}` | Get a slice's full definition, including its output schema | | POST | `/slices/{key}/preview-count` | Count how many calls match your filters — without spending an AI run | | POST | `/slices/{key}/run` | Start an async analysis. Returns `202` with a `runId` for polling | ## Slice Runs Once you start a run, these endpoints let you track progress and retrieve results. Runs are asynchronous — they move through phases (`queued` → `gathering_evidence` → `analyzing_calls` → `synthesizing` → `completed`) and can take seconds to minutes depending on how many calls match. | Method | Path | What it does | | ------ | --------------------------- | --------------------------------------------------------------------------- | | GET | `/slice-runs` | Browse run history, filterable by slice key and status | | GET | `/slice-runs/{id}` | Get a run's current status, progress, and results (when completed) | | GET | `/slice-runs/{id}/evidence` | Get evidence grouped by element — only available for completed runs | | POST | `/evidence/search` | Search evidence across multiple completed runs by element, keyword, or date | ## Posts Posts are the AI-generated outputs from processed calls — either a **summary** (key topics, decisions, next steps) or a **scorecard** (element-by-element rep evaluation). These endpoints let you search, filter, and retrieve them. | Method | Path | What it does | | ------ | ---------------- | ----------------------------------------------------------------- | | POST | `/posts/search` | Search posts by date, call type, rep, CRM deal, company, and more | | GET | `/posts/{id}` | Get a single post with the full structured payload | | GET | `/posts/filters` | Discover which filter keys and values are available for your team | ## Filters Before running a slice or searching posts, you may need to know what values are valid for your filters. These typeahead endpoints let you search for call owners, companies, and CRM entities. | Method | Path | What it does | Scope | | ------ | -------------------------- | -------------------------------------------------------- | ----- | | GET | `/filters/call-owners` | Search call owners by name | Team | | GET | `/filters/company-domains` | Search company domains | Team | | GET | `/filters/crm-deals` | Search CRM deals by name | Org | | GET | `/filters/crm-companies` | Search CRM companies by name | Org | | GET | `/filters/crm-deal-stages` | List CRM deal stages | Org | | GET | `/filters/crm-fields` | List configured CRM custom fields and their valid values | Org | Call owners and company domains are **team-scoped**. CRM data (deals, companies, stages, fields) is **org-scoped** because CRM connections are shared across the organization. ## Reference Data These endpoints return metadata about your team's configuration — useful for understanding what call types and analysis elements exist before building filters or interpreting results. | Method | Path | What it does | | ------ | ------------- | ---------------------------------------------------------------------------------- | | GET | `/call-types` | List active call types (e.g., "Discovery Call", "Demo") with call counts | | GET | `/elements` | List active analysis elements (e.g., "Budget Confirmation", "Competitive Mention") | For full details on how these entities relate to each other, see [Data Models](/api/data-models). # Errors Source: https://docs.oneperfectslice.ai/api/errors Every error the OnePerfectSlice API can return — error envelope shape, HTTP status codes, error code reference, and how to handle each one. When a request fails, the OnePerfectSlice API returns a JSON object with a machine-readable `code` and a human-readable `message`: ```json theme={null} { "error": { "code": "INVALID_FILTERS", "message": "Filter validation failed", "details": { "field": "crmDealStages", "maxAllowed": 100, "validValuesEndpoint": "/api/public/v1/filters/crm-deal-stages" } } } ``` | Field | Type | What it contains | | --------- | -------------- | -------------------------------------------------------------------------------- | | `code` | string | Machine-readable identifier in UPPER\_SNAKE\_CASE | | `message` | string | Plain-English explanation of what went wrong | | `details` | object or null | Additional context when available (e.g., which field failed, valid alternatives) | ## HTTP status codes | Status | What it means | What to do | | ------- | ------------------------------------------------ | ----------------------------------------------------------------------------- | | **200** | Success | — | | **202** | Accepted — an async slice run has started | Poll `GET /slice-runs/{id}` for results | | **400** | Bad request — invalid parameters or filters | Fix the request body; check [filter endpoints](/api/filters) for valid values | | **401** | Unauthorized — missing or invalid API key | Check your key; see [Authentication](/api/authentication) | | **403** | Forbidden — usage limit exceeded | Wait for your billing cycle to reset or contact support | | **404** | Resource not found | Verify the ID and that your key has access to this team's data | | **409** | Conflict — idempotency mismatch or run not ready | Use a new idempotency key, or wait for the run to complete | | **429** | Concurrency limit exceeded | Wait for active runs to finish; see [Rate Limiting](/api/rate-limiting) | | **503** | Timeout or temporary unavailability | Retry with exponential backoff (1s → 2s → 4s) | ## Error code reference ### Authentication (401) | Code | When it happens | | --------------- | ------------------------------------------------ | | `MISSING_TOKEN` | No `Authorization: Bearer` header in the request | | `INVALID_TOKEN` | The key doesn't match any active token | | `TOKEN_REVOKED` | An Org Admin revoked this key | | `TOKEN_EXPIRED` | The key has passed its expiration date | ### Validation (400) | Code | When it happens | | ------------------------- | ------------------------------------------------------ | | `INVALID_REQUEST` | The request body isn't valid JSON | | `INVALID_FILTERS` | One or more filter values are invalid | | `DATE_RANGE_TOO_WIDE` | Date range exceeds the 90-day lookback limit | | `FILTER_ARRAY_TOO_LARGE` | A filter array has more than 100 items | | `INVALID_IDEMPOTENCY_KEY` | Idempotency key is empty or longer than 255 characters | ### Not found (404) | Code | When it happens | | ----------------- | ----------------------------------------------------------- | | `SLICE_NOT_FOUND` | No slice with this key exists for your team | | `RUN_NOT_FOUND` | No run with this ID exists or it belongs to a different org | Resources in a different organization always return 404 — never 403 — so you can't determine whether a resource exists outside your org. ### Conflict (409) | Code | When it happens | | -------------------------- | --------------------------------------------------------- | | `IDEMPOTENCY_KEY_CONFLICT` | You reused an idempotency key with different arguments | | `RUN_NOT_COMPLETED` | You requested evidence for a run that hasn't finished yet | ### Limits (403 / 429) | Code | Status | When it happens | | ---------------------- | ------ | -------------------------------------------------- | | `USAGE_LIMIT_EXCEEDED` | 403 | Your team's monthly AI insights quota is exhausted | | `CONCURRENT_RUN_LIMIT` | 429 | Your team already has 5 active slice runs | ### Timeouts (503) | Code | When it happens | | ----------------------------- | ------------------------------------------------------------------------------------------- | | `PREVIEW_COUNT_TIMEOUT` | Preview count query took too long — try narrowing your filters | | `RUN_LIST_TIMEOUT` | Run history query took too long | | `RUN_LOOKUP_TIMEOUT` | Run detail lookup took too long | | `EVIDENCE_TIMEOUT` | Evidence retrieval took too long | | `POST_SEARCH_TIMEOUT` | Post search took too long — try a shorter date range | | `FILTER_LOOKUP_TIMEOUT` | Filter lookup took too long | | `BILLING_UNAVAILABLE` | Usage gating service is temporarily down | | `IDEMPOTENCY_KEY_IN_PROGRESS` | Another request with this idempotency key is still being processed — retry in a few seconds | # Introduction Source: https://docs.oneperfectslice.ai/api/introduction Programmatic access to your OnePerfectSlice workspace. The OnePerfectSlice API is a REST API that returns JSON. Use it to run AI-powered analysis on your sales conversations, poll for results, search posts, and retrieve evidence. **Base URL** ```text theme={null} https://app.oneperfectslice.ai/api/public/v1 ``` All requests require a Bearer token — see [Authentication](/api/authentication). ## Response envelope All successful responses wrap data in a standard envelope: ```json theme={null} { "data": { ... }, "meta": { "pagination": { "totalCount": 142, "hasMore": true, "limit": 20, "offset": 0 } } } ``` The `meta` field is only present on paginated endpoints. Non-paginated endpoints return `{ "data": { ... } }` only. ## Error envelope Error responses use a separate structure with a machine-readable `code` (UPPER\_SNAKE\_CASE) and a human-readable `message`: ```json theme={null} { "error": { "code": "INVALID_TOKEN", "message": "Token is invalid", "details": {} } } ``` The `details` field is optional and provides additional context for certain errors. See [Errors](/api/errors) for the full list of error codes. ## Endpoint groups | Group | Description | | -------------------------------- | --------------------------------------------------------- | | [Slices](/api/slices) | List, inspect, preview, and run AI analysis templates | | [Slice Runs](/api/slice-runs) | Poll run status, retrieve results and evidence | | [Posts](/api/posts) | Search and retrieve AI-generated summaries and scorecards | | [Reference Data](/api/reference) | Call types and elements for your team | | [Filters](/api/filters) | Typeahead endpoints for building filter UIs | ## Quickstart For a step-by-step walkthrough, see the [Quickstart](/quickstart). # Pagination Source: https://docs.oneperfectslice.ai/api/pagination The OnePerfectSlice API uses offset-based pagination. Pass limit and offset as query parameters to control which page of results you receive. ## Request parameters | Parameter | Type | Default | Range | What it does | | --------- | ------- | ------- | ----- | ------------------------------------------------- | | `limit` | integer | 20 | 1–100 | How many results to return per page | | `offset` | integer | 0 | 0+ | How many results to skip before starting the page | ## Response metadata Every paginated response includes a `meta.pagination` object that tells you where you are in the result set: ```json theme={null} { "data": [ ... ], "meta": { "pagination": { "totalCount": 142, "hasMore": true, "limit": 20, "offset": 0 } } } ``` | Field | Type | What it tells you | | ------------ | ------- | ------------------------------------------- | | `totalCount` | integer | Total results across all pages | | `hasMore` | boolean | Whether there are more pages after this one | | `limit` | integer | Page size (echoed from your request) | | `offset` | integer | Current position in the result set | ## Which endpoints are paginated? | Endpoint | What it returns | | -------------------- | ---------------------------------------------- | | `GET /slice-runs` | Run history for your team | | `POST /posts/search` | Summaries and scorecards matching your filters | All other endpoints return the full result set in a single response. ## Fetching all pages Use the `hasMore` field to loop through pages. Increment `offset` by `limit` each time until `hasMore` is `false`. ```bash cURL theme={null} # First page curl "https://app.oneperfectslice.ai/api/public/v1/slice-runs?limit=20&offset=0" \ -H "Authorization: Bearer sk_your_api_key" # Second page curl "https://app.oneperfectslice.ai/api/public/v1/slice-runs?limit=20&offset=20" \ -H "Authorization: Bearer sk_your_api_key" ``` ```python Python theme={null} import requests headers = {"Authorization": "Bearer sk_your_api_key"} base = "https://app.oneperfectslice.ai/api/public/v1/slice-runs" all_runs = [] offset = 0 limit = 20 while True: resp = requests.get(base, headers=headers, params={"limit": limit, "offset": offset}) page = resp.json() all_runs.extend(page["data"]) if not page["meta"]["pagination"]["hasMore"]: break offset += limit print(f"Fetched {len(all_runs)} runs") ``` ```typescript TypeScript theme={null} const headers = { Authorization: "Bearer sk_your_api_key" }; const base = "https://app.oneperfectslice.ai/api/public/v1/slice-runs"; const allRuns: any[] = []; let offset = 0; const limit = 20; while (true) { const resp = await fetch(`${base}?limit=${limit}&offset=${offset}`, { headers }); const page = await resp.json(); allRuns.push(...page.data); if (!page.meta.pagination.hasMore) break; offset += limit; } console.log(`Fetched ${allRuns.length} runs`); ``` # Rate Limiting Source: https://docs.oneperfectslice.ai/api/rate-limiting Understand rate limits, concurrency limits, and timeout behavior in the OnePerfectSlice API. Learn how to handle 429 and 503 responses. The OnePerfectSlice API enforces concurrency limits on slice runs and has server-side timeouts on long-running queries. Per-request rate limits are not currently enforced but may be introduced in the future. ## Concurrency limits Each team can run up to **5 slice runs at the same time**. Attempting to start a 6th returns a `429` error: ```json theme={null} { "error": { "code": "CONCURRENT_RUN_LIMIT", "message": "Maximum of 5 concurrent slice runs per team. Wait for existing runs to complete." } } ``` To avoid hitting this limit, poll your active runs with `GET /slice-runs?status=running` before starting new ones. ## Per-request rate limits The API does not currently enforce per-request rate limits. When they are introduced, this page will be updated with: * Per-token request limits and window duration * Rate limit headers (`X-RateLimit-Limit`, `X-RateLimit-Remaining`, `X-RateLimit-Reset`) * Recommended backoff strategies Your client should be prepared to handle `429 Too Many Requests` responses with a retry-after delay. ## Timeouts Some endpoints involve database queries or downstream processing that can time out under load. When this happens, the API returns `503` with a specific error code: | Error code | Endpoint | What happened | | ----------------------- | ---------------- | -------------------------------- | | `PREVIEW_COUNT_TIMEOUT` | Preview count | Filter count query took too long | | `RUN_LIST_TIMEOUT` | List runs | Run history query took too long | | `RUN_LOOKUP_TIMEOUT` | Get run | Run detail lookup took too long | | `EVIDENCE_TIMEOUT` | Get evidence | Evidence retrieval took too long | | `POST_SEARCH_TIMEOUT` | Search posts | Post search took too long | | `FILTER_LOOKUP_TIMEOUT` | Filter endpoints | Filter lookup took too long | **How to handle timeouts:** Retry with exponential backoff — wait 1s, then 2s, then 4s. If the timeout persists, try narrowing your filters or reducing the date range. # Overview Source: https://docs.oneperfectslice.ai/index OnePerfectSlice is a context-as-a-service platform for B2B GTM teams. It builds structured, evidence-backed context compiled from your data and delivers it to AI agents, workflows, and teams — so everyone can make smarter decisions, faster, with fewer hand-offs. ## Who it's for * **Sales & Marketing Operations** — make every team, agent, and workflow customer-aware * **GTM Engineers and Marketing Engineers** — build context-aware automations and AI workflows without stitching together fragmented data * **Demand Gen Teams** — improve targeting and campaign performance with timely signals from across your stack * **Product Marketing Teams** — uncover customer, market, and competitive intelligence to sharpen positioning and messaging * **Sales Enablement Teams** — coach reps with evidence from real conversations and reinforce winning behaviors at scale ## What you can do | Capability | Description | | ------------------------ | ----------------------------------------------------------------------------------------------------------------------- | | **Feed your AI agents** | Agents get real-time, compiled context from across your stack via MCP. No copy-pasting, no manual context-building. | | **Power your workflows** | Connect OnePerfectSlice to Zapier, n8n, Make, Clay, or your own code via the public API. | | **Equip your team** | Summaries, scorecards, and evidence show up where your people work — decisions happen faster, with less back-and-forth. | ## Go deeper AI analysis templates that surface structured insights tied to business goal and role. Every call is auto-classified by type, then summarized with a GTM-focused template — enabling consistent comparison and pattern matching across your team. Each call type has a structured scorecard template — reps are evaluated against consistent criteria so you can spot trends at scale. Connect AI agents to compiled context from across your stack — ask questions in natural language. 20 REST endpoints for programmatic access to slices, runs, posts, and evidence. Gong, Fireflies, Fathom, HubSpot, Notion — and what's coming next. # Fathom Source: https://docs.oneperfectslice.ai/integrations/fathom Connect Fathom to OnePerfectSlice to automatically import your call recordings for AI analysis, summaries, and scorecards. ## Connect Fathom Navigate to **Org Settings → Integrations**. You'll need Org Admin permissions. A modal will open asking for your Fathom API key. Paste your Fathom API key. You can find it in your [Fathom account settings](https://fathom.video/support). Once connected, OnePerfectSlice imports your recent calls and begins syncing new ones automatically every 15 minutes. ## After connecting * Existing calls are backfilled automatically (the number depends on your plan) * New calls sync every **15 minutes** * Each call is [automatically classified](/product/call-types), then summaries and scorecards are generated ## Managing your connection From the Fathom integration card in Org Settings, you can: | Action | What it does | | -------------- | ---------------------------------------------------- | | **Pause** | Stop syncing new calls (existing data is preserved) | | **Resume** | Restart syncing | | **Reconnect** | Update your API key without losing existing data | | **Disconnect** | Remove the connection (existing calls are preserved) | # Fireflies Source: https://docs.oneperfectslice.ai/integrations/fireflies Connect Fireflies to OnePerfectSlice to automatically import your call recordings for AI analysis, summaries, and scorecards. ## Connect Fireflies Navigate to **Org Settings → Integrations**. You'll need Org Admin permissions. A modal will open asking for your Fireflies API key. Paste your Fireflies API key. You can find it in your [Fireflies account settings](https://docs.fireflies.ai/). Once connected, OnePerfectSlice imports your recent calls and begins syncing new ones automatically every 15 minutes. ## After connecting * Existing calls are backfilled automatically (the number depends on your plan) * New calls sync every **15 minutes** * Each call is [automatically classified](/product/call-types), then summaries and scorecards are generated ## Managing your connection From the Fireflies integration card in Org Settings, you can: | Action | What it does | | -------------- | ---------------------------------------------------- | | **Pause** | Stop syncing new calls (existing data is preserved) | | **Resume** | Restart syncing | | **Reconnect** | Update your API key without losing existing data | | **Disconnect** | Remove the connection (existing calls are preserved) | # Gong Source: https://docs.oneperfectslice.ai/integrations/gong Connect Gong to OnePerfectSlice to automatically import your call recordings for AI analysis, summaries, and scorecards. ## Connect Gong Navigate to **Org Settings → Integrations**. You'll need Org Admin permissions. A modal will open asking for your Gong API credentials. You'll need three things from Gong: | Field | Where to find it | | --------------------- | ---------------------------------------------------------------------- | | **API Base URL** | Your regional Gong API endpoint (e.g., `https://us-12345.api.gong.io`) | | **Access Key** | From your Gong API settings | | **Access Key Secret** | From your Gong API settings | Not sure where to find these? See [Gong's API access guide](https://help.gong.io/docs/receive-access-to-the-api). Once connected, OnePerfectSlice imports your recent calls and begins syncing new ones automatically every 15 minutes. ## After connecting * Existing calls are backfilled automatically (the number depends on your plan) * New calls sync every **15 minutes** * Each call is [automatically classified](/product/call-types), then summaries and scorecards are generated ## Managing your connection From the Gong integration card in Org Settings, you can: | Action | What it does | | -------------- | ---------------------------------------------------- | | **Pause** | Stop syncing new calls (existing data is preserved) | | **Resume** | Restart syncing | | **Reconnect** | Update your credentials without losing existing data | | **Disconnect** | Remove the connection (existing calls are preserved) | # HubSpot Source: https://docs.oneperfectslice.ai/integrations/hubspot Connect HubSpot to OnePerfectSlice to enrich your structured context with deal, company, and contact data from your CRM. ## What gets synced When you connect HubSpot, OnePerfectSlice syncs three entity types: | Entity | Data synced | | ------------- | --------------------------------------------------------------- | | **Deals** | Name, stage, pipeline, amount, close date, owner, custom fields | | **Companies** | Name, domain, lifecycle stage, owner, custom fields | | **Contacts** | Email, name, lifecycle stage, owner, custom fields | OnePerfectSlice associates CRM records with calls by matching participant email addresses. ## Connect HubSpot Navigate to **Org Settings → Integrations**. You'll need Org Admin permissions. Click **Connect** on the HubSpot card. An OAuth popup will appear — authorize OnePerfectSlice to access your HubSpot account. After connecting, click **Settings** to configure: * **Sync Settings** — toggle Deals, Companies, and Contacts on or off * **Field Settings** — select which fields to sync (up to 10 custom fields per entity) ## Sync schedule * CRM data syncs automatically **every 4 hours** * You can trigger a manual resync from the integration settings (rate-limited to once per hour) * OAuth tokens are refreshed automatically ## How CRM data enriches your context Once connected, CRM data becomes part of the structured context OnePerfectSlice delivers: * **Filtering** — filter summaries, scorecards, and slice runs by CRM deal, company, stage, or amount * **Enrichment** — CRM context is included when slices analyze calls, producing richer and more grounded evidence * **Custom fields** — map CRM fields to create custom filters (e.g., filter by industry, deal size tier, or any custom property) ## Filter endpoints The API provides endpoints for searching CRM data programmatically: * `/filters/crm-deals` — search deals by name * `/filters/crm-companies` — search companies by name * `/filters/crm-deal-stages` — list available deal stages * `/filters/crm-fields` — list configured custom fields and their valid values CRM data is **org-scoped** because the CRM connection is shared across the organization. Summaries, scorecards, and calls remain **team-scoped**. ## Disconnecting You can disconnect HubSpot at any time from Org Settings. Disconnecting clears the OAuth tokens but retains previously synced CRM data. # Integrations Source: https://docs.oneperfectslice.ai/integrations/index OnePerfectSlice connects to your call recording platform, CRM, and productivity tools. See what's available and what's coming next. OnePerfectSlice integrates with the tools you already use — pull in call recordings, sync with your CRM, and push insights to your productivity stack. ## Call Recording Connect your call recording platform to automatically ingest transcripts for analysis. | Integration | Status | Setup | | ------------- | ----------- | -------------------------------------- | | **Gong** | Available | [Setup guide](/integrations/gong) | | **Fireflies** | Available | [Setup guide](/integrations/fireflies) | | **Fathom** | Available | [Setup guide](/integrations/fathom) | | **Avoma** | Coming soon | — | ## CRM Sync deal, company, and contact data to enrich your analyses with CRM context. | Integration | Status | Setup | | -------------- | ----------- | ------------------------------------ | | **HubSpot** | Available | [Setup guide](/integrations/hubspot) | | **Salesforce** | Coming soon | — | ## Productivity Tools Push insights, scorecards, and summaries to the tools where your team works. | Integration | Status | Setup | | ----------- | ----------- | ----------------------------------- | | **Notion** | Available | [Setup guide](/integrations/notion) | | **Slack** | Coming soon | — | | **Asana** | Coming soon | — | | **Monday** | Coming soon | — | Need an integration we don't support yet? [Let us know](mailto:support@oneperfectslice.ai) — we prioritize based on customer demand. # Notion Source: https://docs.oneperfectslice.ai/integrations/notion Connect Notion to OnePerfectSlice to automatically push call summaries, scorecards, and slice results to your team's workspace. ## Connect Notion Navigate to **Org Settings → Integrations**. You'll need Org Admin permissions. An OAuth popup will open — authorize OnePerfectSlice to access your Notion workspace. After connecting, a settings modal opens. Search for and select the Notion page or database where you want content to land. Enable the sync types you want: | Sync type | What gets pushed to Notion | | ------------------ | ------------------------------------------------ | | **Call Reviews** | Summaries and scorecards for each processed call | | **Insight Briefs** | AI-generated insight briefs | | **Slice Runs** | Results from completed slice analyses | Each toggle can be turned on or off independently. ## After connecting * Content syncs automatically when new results are generated * Each sync type tracks its own "last synced" timestamp * You can change the destination page or toggle sync types at any time from the Notion settings card ## Managing your connection From the Notion integration card in Org Settings, you can: | Action | What it does | | --------------------- | ----------------------------------------------------------- | | **Edit destination** | Change which Notion page receives synced content | | **Toggle sync types** | Turn individual sync types on or off | | **Reconnect** | Re-authorize if your Notion token was revoked | | **Disconnect** | Remove the connection (existing Notion pages are preserved) | # ChatGPT Source: https://docs.oneperfectslice.ai/mcp/chatgpt Connect ChatGPT to the OnePerfectSlice remote MCP server as a custom connector with OAuth — no terminal required. ChatGPT connects to the hosted OnePerfectSlice server as a **custom connector**. Sign-in happens in your browser via OAuth — there's nothing to install. Adding a custom MCP connector requires **Developer Mode**, available on paid plans (Plus, Pro, Team, Enterprise, Edu) and currently in beta. On Team/Enterprise/Edu, an admin may need to enable connectors for the workspace first. ## Connect In ChatGPT, go to **Settings → Apps & Connectors → Advanced settings** and turn on **Developer Mode**. This unlocks adding custom MCP servers with full tool access. Back in **Settings → Apps & Connectors**, choose **Create / Add custom connector** and enter: * **Name:** `OnePerfectSlice` * **MCP server URL:** `https://mcp.oneperfectslice.ai/mcp` When prompted, sign in to OnePerfectSlice and **select the team** whose data this connection should access. ChatGPT completes the OAuth flow and connects. Enable the `OnePerfectSlice` connector in a chat (via the **+ / apps & connectors** menu in the composer), then ask it to list your slices. ## Test it In a chat with the connector enabled, try: ``` Use OnePerfectSlice to list available slices ``` You should get back a list of your team's slices. ## Troubleshooting | Problem | Solution | | ---------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------- | | No "custom connector" option | Turn on **Developer Mode** under Settings → Apps & Connectors → Advanced settings (Team/Enterprise/Edu may need an admin to enable connectors) | | Connector won't authorize | Re-run the sign-in and complete both login and team selection | | Wrong team's data | Remove the connector, re-add it, and select a different team | | Tools not available in chat | Enable the connector from the chat's tools/connectors menu | ## Switching teams Each connection is scoped to one team. To switch, remove the connector and add it again, selecting the new team during sign-in. # Claude Code Source: https://docs.oneperfectslice.ai/mcp/claude-code Connect Claude Code to the OnePerfectSlice remote MCP server — add the URL, authenticate with OAuth, verify, and troubleshoot. The OnePerfectSlice MCP server is hosted — there's nothing to install. You add the URL once and authenticate through your browser. ## Connect ```bash theme={null} claude mcp add --transport http ops https://mcp.oneperfectslice.ai/mcp ``` Add `--scope user` to make `ops` available across all your projects: ```bash theme={null} claude mcp add --transport http ops --scope user https://mcp.oneperfectslice.ai/mcp ``` In Claude Code, open the MCP panel: ``` /mcp ``` Select **ops**, choose **Authenticate**, and your browser will open. Log in to OnePerfectSlice and **select the team** whose data this connection should access. Claude Code receives the token and reconnects automatically. Run `/mcp` again — `ops` should show as **connected** with 9 tools. ## Test it In Claude Code, try: ``` Use ops to list available slices ``` You should see a tool call to `list_slices` and get back a list of your team's slices. ## Troubleshooting | Problem | Solution | | ---------------------------- | ---------------------------------------------------------------------------------- | | `ops` not showing in `/mcp` | Re-run the `claude mcp add` command, then restart Claude Code | | Stuck "needs authentication" | Run `/mcp`, select `ops`, and choose **Authenticate** to complete the browser flow | | Browser flow didn't finish | Make sure you completed login **and** team selection; then retry from `/mcp` | | Wrong team's data | Reconnect from `/mcp` and select a different team during authentication | | Session expired | Tokens refresh automatically; if a call fails, re-authenticate from `/mcp` | ## Switching teams Each connection is scoped to one team. To point `ops` at a different team, run `/mcp`, choose **ops**, re-authenticate, and select the new team. # Claude Desktop Source: https://docs.oneperfectslice.ai/mcp/claude-desktop Connect Claude Desktop to the OnePerfectSlice remote MCP server using a custom connector and OAuth — no config files, no API keys. The OnePerfectSlice MCP server is hosted, so there's no config file to edit and no key to paste. You add it as a **custom connector** and sign in through your browser. ## Connect In Claude Desktop, go to **Settings → Connectors**. Click **Add custom connector**, then enter: * **Name:** `OnePerfectSlice` * **URL:** `https://mcp.oneperfectslice.ai/mcp` Save the connector. Click **Connect** on the new connector. Your browser opens to OnePerfectSlice — log in and **select the team** whose data this connection should access. Claude Desktop receives the token and connects automatically. Back in **Settings → Connectors**, confirm `OnePerfectSlice` shows as **connected**. ## Test it Start a new conversation and try: ``` Use OnePerfectSlice to list available slices ``` You should see a tool call to `list_slices` and get back a list of your team's slices. ## Troubleshooting | Problem | Solution | | -------------------------- | ---------------------------------------------------------------------- | | Connector won't connect | Confirm the URL is exactly `https://mcp.oneperfectslice.ai/mcp` | | Browser flow didn't finish | Complete login **and** team selection, then click **Connect** again | | Wrong team's data | Disconnect the connector, reconnect, and select a different team | | Tools don't appear | Fully quit Claude Desktop (Cmd+Q on macOS) and relaunch | | Session expired | Tokens refresh automatically; if a call fails, reconnect the connector | ## Switching teams Each connection is scoped to one team. To switch, disconnect the connector under **Settings → Connectors**, reconnect, and select the new team during sign-in. # Codex Source: https://docs.oneperfectslice.ai/mcp/codex Connect OpenAI Codex to the OnePerfectSlice remote MCP server over streamable HTTP with OAuth login. Codex connects to the hosted OnePerfectSlice server over streamable HTTP. You register the URL in your Codex config, then authenticate with `codex mcp login`. ## Connect Add the server to `~/.codex/config.toml`: ```toml theme={null} [mcp_servers.ops] url = "https://mcp.oneperfectslice.ai/mcp" ``` OAuth login is supported only for streamable HTTP servers — the `url` field is what enables it. Don't set a `command` for `ops`. Run the OAuth login flow: ```bash theme={null} codex mcp login ops ``` Your browser opens to OnePerfectSlice — log in and **select the team** whose data this connection should access. Codex stores the token in its managed encrypted credential storage. ```bash theme={null} codex mcp list ``` You should see `ops` listed and authenticated. ## Test it In Codex, try: ``` Use ops to list available slices ``` You should see a tool call to `list_slices` and get back a list of your team's slices. ## Troubleshooting | Problem | Solution | | ----------------------------- | ------------------------------------------------------------------------------------------- | | `ops` not listed | Check the `[mcp_servers.ops]` block in `~/.codex/config.toml` for typos, then restart Codex | | Login didn't complete | Re-run `codex mcp login ops` and finish both login and team selection in the browser | | Wrong team's data | Run `codex mcp login ops` again and select a different team | | Tool calls fail after a while | Re-run `codex mcp login ops` to refresh credentials | ## Switching teams Each connection is scoped to one team. To switch, run `codex mcp login ops` again and select the new team during sign-in. # Cursor Source: https://docs.oneperfectslice.ai/mcp/cursor Connect Cursor to the OnePerfectSlice remote MCP server with OAuth — add the URL in MCP settings. Cursor connects to the hosted OnePerfectSlice server over HTTP and signs in with OAuth. Nothing to install. ## Connect In Cursor, open **Settings → Tools & Integrations** (the MCP section) and choose **Add custom MCP / New MCP server**. This opens the `mcp.json` config — global at `~/.cursor/mcp.json`, or per-project at `.cursor/mcp.json`. Add the OnePerfectSlice URL. A remote server needs only `url` — no `type` or `transport` field: ```json theme={null} { "mcpServers": { "ops": { "url": "https://mcp.oneperfectslice.ai/mcp" } } } ``` Cursor detects that the server requires OAuth and shows a **Sign in / Authenticate** prompt for `ops`. Click it, log in to OnePerfectSlice, and **select the team** whose data this connection should access. In **Settings → MCP**, confirm `ops` is enabled and its tools are listed. ## Test it In Cursor's chat (Agent mode), try: ``` Use ops to list available slices ``` You should see a `list_slices` tool call and get back your team's slices. ## Troubleshooting | Problem | Solution | | -------------------- | --------------------------------------------------------------------------- | | `ops` shows no tools | Click **Authenticate** next to `ops` in Settings → MCP and complete sign-in | | Authentication loops | Confirm the URL is exactly `https://mcp.oneperfectslice.ai/mcp`, then retry | | Wrong team's data | Remove `ops` from `mcp.json`, re-add it, and select a different team | | Tools not used | Make sure you're in Agent mode and `ops` is enabled | ## Switching teams Each connection is scoped to one team. To switch, remove and re-add the `ops` entry, then select the new team during sign-in. # Other clients Source: https://docs.oneperfectslice.ai/mcp/other-clients Connect any MCP-compatible client to the OnePerfectSlice remote server — one URL, OAuth, no install. The OnePerfectSlice server is a standard **remote MCP server with OAuth**, so it works with any spec-compliant client. There's only one thing to configure everywhere: ``` https://mcp.oneperfectslice.ai/mcp ``` Add it as a **streamable HTTP** (also called "remote" or "URL") MCP server. On first connect, the client opens your browser to log in and **select a team**; tokens refresh automatically after that. ## Where to add the URL | Client | Where to add it | | ---------------------------------- | ------------------------------------------------------------------------- | | **Windsurf** | Settings → Cascade → MCP servers → add an HTTP/SSE server with the URL | | **Cline** (VS Code) | MCP Servers panel → Remote Servers → add the URL | | **Goose** | Add an extension of type *Streamable HTTP* with the URL | | **Zed** | `settings.json` → `context_servers` → add an HTTP MCP server with the URL | | **JetBrains AI Assistant / Junie** | Settings → Tools → MCP → add a remote server with the URL | | **LM Studio** | Program → Edit `mcp.json` → add the server URL | | **Continue** | Add an MCP server block referencing the URL | | **Warp** | Settings → AI → MCP servers → add the URL | | **Raycast AI** | AI → MCP servers → add the URL | Exact menu names vary by client and version. Look for "MCP servers," "Connectors," or "Tools," and choose the **remote / HTTP / URL** option (not stdio or "local command"). If a client offers only stdio, it can't use this server directly. ## What you get Once connected, all [9 tools](/mcp/tools) are available — `list_slices`, `run_slice`, `search_posts`, `search_evidence`, and the rest. Test any client with: ``` Use OnePerfectSlice to list available slices ``` ## Don't see your client? Any client that supports remote MCP servers with OAuth will work. If yours expects a manual token instead of a browser sign-in, it isn't compatible with this server's OAuth flow — reach out and we'll help. # Getting Started Source: https://docs.oneperfectslice.ai/mcp/overview The OnePerfectSlice MCP server connects your AI agents to structured context from your calls and CRM. Analyze patterns, search posts, and retrieve evidence — all through natural language in Claude Code, Claude Desktop, or Codex. The OnePerfectSlice MCP server is **hosted** — there's nothing to download, install, or update. Point your client at one URL and sign in through your browser: ``` https://mcp.oneperfectslice.ai/mcp ``` ## Authentication The server uses **OAuth 2.0** — no API keys to copy or store. The first time you connect from any client: 1. Your browser opens to OnePerfectSlice 2. You log in (or you're already signed in) 3. You **select the team** whose data this connection should access 4. The client receives a token and reconnects automatically Tokens refresh on their own, so you authenticate once per client. To switch teams, reconnect and pick a different team. ## Set up your client Setup takes under a minute — add the URL, then authenticate. Pick your client: One CLI command. Best for developers in the terminal. Add a custom connector. Best for non-technical users. Add a connector in settings. No terminal required. Add the URL in MCP settings. Connect through GitHub Copilot's MCP support. Configure and run `codex mcp login`. Windsurf, Cline, Goose, Zed, LM Studio and more — the same URL works anywhere. ## How it works The MCP server sits on top of the [OnePerfectSlice REST API](/api/introduction) and exposes 9 tools designed for agent interaction. Key differences from calling the API directly: * **Composite operations** — `run_slice` starts a run, polls for completion, and fetches evidence in one call, not three. * **Flat parameters** — no nested objects. Every filter is a top-level argument the agent can set directly. * **Guided errors** — when something goes wrong, error messages tell the agent exactly which tool to call next. ## Available tools | Tool | What it does | | ------------------------------------------------------- | --------------------------------------------------------------- | | [`list_slices`](/mcp/tools#list_slices) | Discover available analyses or get details for a specific slice | | [`preview_slice_count`](/mcp/tools#preview_slice_count) | Count matched calls before running (free, no AI cost) | | [`run_slice`](/mcp/tools#run_slice) | Execute an analysis and get results + evidence | | [`get_run`](/mcp/tools#get_run) | Retrieve results from a past or in-progress run | | [`list_runs`](/mcp/tools#list_runs) | Browse run history with optional filtering | | [`search_posts`](/mcp/tools#search_posts) | Find summaries and scorecards matching your criteria | | [`get_post`](/mcp/tools#get_post) | Get a single post with full structured content | | [`search_evidence`](/mcp/tools#search_evidence) | Search evidence quotes across all completed runs | | [`get_filter_options`](/mcp/tools#get_filter_options) | Discover valid filter values (call owners, companies, CRM data) | See [Tools](/mcp/tools) for the full parameter reference with examples. # Tools Source: https://docs.oneperfectslice.ai/mcp/tools Complete reference for all 9 tools available in the OnePerfectSlice MCP server — parameters, return values, and example prompts for each. The OnePerfectSlice MCP server exposes 9 tools. Each tool maps to one or more API endpoints but is designed for conversational use — flat parameters, composite operations, and error messages that guide the agent to the right next step. ## list\_slices Discover what analyses are available, or get full details for a specific slice. | Parameter | Type | Required | Description | | ----------- | ------ | -------- | ---------------------------------------------------------------------------------------- | | `slice_key` | string | No | If provided, returns full detail with output schema. Otherwise returns the full catalog. | **Example prompts:** * "What slices do I have access to?" * "Show me the details for the high\_intent\_deals slice" *** ## preview\_slice\_count Check how many calls match your filters — without starting a run or using AI credits. | Parameter | Type | Required | Description | | ------------------- | ---------- | -------- | ------------------------------------------- | | `slice_key` | string | Yes | Which slice to preview | | `date_from` | string | Yes | Start date (ISO format, e.g., `2026-03-26`) | | `date_to` | string | Yes | End date | | `call_types` | string\[] | No | Filter by call type keys | | `call_owner_emails` | string\[] | No | Filter by rep emails | | `company_domains` | string\[] | No | Filter by company domains | | `crm_deal_stages` | string\[] | No | Filter by CRM deal stages | | `crm_deal_ids` | integer\[] | No | Filter by CRM deal IDs | | `crm_company_ids` | integer\[] | No | Filter by CRM company IDs | | `crm_amount_min` | number | No | Minimum deal amount | | `crm_amount_max` | number | No | Maximum deal amount | | `crm_field_filters` | string\[] | No | Custom CRM filters (`"field=value"` syntax) | **Example prompts:** * "How many discovery calls happened last month?" * "Preview how many calls match the high\_intent\_deals slice for Q1" *** ## run\_slice Run an AI analysis and get results. Waits up to 30 seconds for completion — if the run takes longer, returns a run ID for polling with `get_run`. | Parameter | Type | Required | Description | | ------------------- | ---------- | -------- | ------------------------------------------- | | `slice_key` | string | Yes | Which slice to run | | `date_from` | string | Yes | Start datetime (ISO format) | | `date_to` | string | Yes | End datetime | | `call_types` | string\[] | No | Filter by call type keys | | `call_owner_emails` | string\[] | No | Filter by rep emails | | `company_domains` | string\[] | No | Filter by company domains | | `crm_deal_stages` | string\[] | No | Filter by CRM deal stages | | `crm_deal_ids` | integer\[] | No | Filter by CRM deal IDs | | `crm_company_ids` | integer\[] | No | Filter by CRM company IDs | | `crm_amount_min` | number | No | Minimum deal amount | | `crm_amount_max` | number | No | Maximum deal amount | | `crm_field_filters` | string\[] | No | Custom CRM filters (`"field=value"` syntax) | **What happens behind the scenes:** 1. Starts an async run via the API 2. Polls for up to 30 seconds 3. If completed, automatically fetches evidence and returns everything together 4. If still running, returns the run ID with a progress message **Example prompts:** * "Run the high intent deals analysis on last month's calls" * "Analyze discovery calls from the past 2 weeks for the enterprise team" *** ## get\_run Retrieve results from a past run, or check progress on a running one. By default it returns the synthesized result without pulling all evidence — drill into evidence only when you need it. | Parameter | Type | Required | Description | | ------------------- | --------- | -------- | ------------------------------------------------------------------------ | | `run_id` | integer | Yes | The run ID (returned by `run_slice` or `list_runs`) | | `summary_only` | boolean | No | Return just the synthesis — a convenience alias for `include=["result"]` | | `include` | string\[] | No | What to return: any of `result`, `evidence` | | `element_keys` | string\[] | No | When fetching evidence, limit to these analysis elements | | `call_owner_emails` | string\[] | No | When fetching evidence, limit to these reps | **Example prompts:** * "Get the summary for run 12345" * "Show me the budget-objection evidence from run 12345" * "Check if my last analysis is done yet" *** ## list\_runs Browse your team's run history with optional filtering. | Parameter | Type | Required | Description | | ------------------- | --------- | -------- | ---------------------------------------------- | | `slice_key` | string | No | Filter by slice | | `status` | string | No | `pending`, `running`, `completed`, or `failed` | | `call_owner_emails` | string\[] | No | Filter to runs scoped to these reps | | `limit` | integer | No | Number of results (1–100) | | `offset` | integer | No | Pagination offset | **Example prompts:** * "Show me recent completed runs" * "What runs are currently active?" * "List the last 5 high\_intent\_deals runs" * "Find Jane's recent runs I can reuse" *** ## search\_posts Find AI-generated summaries and scorecards matching your criteria. Returns compact rows by default — each carries a `fileId` so summary and scorecard rows for the same call can be deduplicated. Use `template_type` to narrow to one artifact, or `detail=true` / [`get_post`](#get_post) for full content. | Parameter | Type | Required | Description | | ------------------- | ---------- | -------- | ------------------------------------------------------ | | `template_type` | string | No | `summary` or `scorecard` | | `call_type_ids` | integer\[] | No | Filter by call type IDs | | `date_from` | string | No | Start datetime | | `date_to` | string | No | End datetime | | `call_owner_emails` | string\[] | No | Filter by rep emails | | `company_domains` | string\[] | No | Filter by company domains | | `crm_deal_stages` | string\[] | No | Filter by CRM deal stages | | `crm_deal_ids` | integer\[] | No | Filter by CRM deal IDs | | `crm_company_ids` | integer\[] | No | Filter by CRM company IDs | | `crm_amount_min` | number | No | Minimum deal amount | | `crm_amount_max` | number | No | Maximum deal amount | | `crm_field_filters` | string\[] | No | Custom CRM filters | | `detail` | boolean | No | Return full structured content instead of compact rows | | `fields` | string\[] | No | Limit the returned fields | | `limit` | integer | No | Results per page (1–100) | | `offset` | integer | No | Pagination offset | **Example prompts:** * "Find scorecards for Jane's calls last week" * "Search for summaries mentioning Acme Corp" * "Get the most recent 10 discovery call summaries" *** ## get\_post Retrieve a single post with full structured content. | Parameter | Type | Required | Description | | --------- | ------- | -------- | ----------- | | `post_id` | integer | Yes | The post ID | **Example prompts:** * "Show me post 456" * "Get the full scorecard for that call" *** ## search\_evidence Search for evidence quotes across all completed runs. Useful for finding patterns or specific mentions across multiple analyses. | Parameter | Type | Required | Description | | ------------------- | --------- | -------- | ---------------------------------------------------- | | `element_keys` | string\[] | Yes | Which elements to search (e.g., `["intent_signal"]`) | | `query` | string | No | Free-text search across quotes | | `call_types` | string\[] | No | Filter by call type keys | | `call_owner_emails` | string\[] | No | Filter by rep emails | | `date_from` | string | No | Start date | | `date_to` | string | No | End date | | `limit` | integer | No | Max results (1–100) | **Example prompts:** * "Find evidence of budget objections in discovery calls" * "Search for competitive mentions from the last quarter" * "Show me intent signals from Jane's calls last month" `search_evidence` filters by element, rep, call type, and date — not by company. To scope evidence to one account, use [`search_posts`](#search_posts) (which supports `company_domains`) or run a slice filtered to that company. *** ## get\_filter\_options Discover valid values for any filter type. Call this before filtering when you're not sure what values are available. | Parameter | Type | Required | Description | | ------------- | ------ | -------- | --------------------------------------------- | | `filter_type` | string | Yes | See table below | | `query` | string | No | Prefix search (for typeahead-capable filters) | **Available filter types:** | Value | What it returns | Supports `query` | | ----------------- | -------------------------------- | ---------------- | | `call_types` | Call type keys and labels | No | | `elements` | Analysis element keys and labels | No | | `call_owners` | Rep names and emails | Yes | | `company_domains` | Company domains | Yes | | `crm_deals` | CRM deal names | Yes | | `crm_companies` | CRM company names | Yes | | `crm_deal_stages` | Pipeline stages | No | | `crm_fields` | Custom CRM field definitions | No | | `post_filters` | Available post filter keys | No | **Example prompts:** * "What call types are available?" * "Search for call owners named Jane" * "What CRM deal stages exist?" * "Show me available CRM custom fields" # VS Code Source: https://docs.oneperfectslice.ai/mcp/vscode Connect VS Code (GitHub Copilot) to the OnePerfectSlice remote MCP server over HTTP with OAuth. VS Code connects to the hosted OnePerfectSlice server through **GitHub Copilot's** MCP support. Nothing to install beyond Copilot. MCP servers are used by Copilot's **Agent mode** in the Chat view. Make sure GitHub Copilot is installed and you're signed in. ## Connect Open the Command Palette (`Cmd/Ctrl+Shift+P`) and run **MCP: Add Server**. Choose **HTTP** and enter: * **URL:** `https://mcp.oneperfectslice.ai/mcp` * **Name:** `ops` This writes an entry to your `mcp.json` (workspace `.vscode/mcp.json` or your user profile): ```json theme={null} { "servers": { "ops": { "type": "http", "url": "https://mcp.oneperfectslice.ai/mcp" } } } ``` Start the server when prompted (or run **MCP: List Servers → ops → Start**). VS Code opens a browser for OAuth — log in to OnePerfectSlice and **select the team** whose data this connection should access. Approve the authorization back in VS Code if asked. Run **MCP: List Servers** and confirm `ops` is running with its tools available. In the Chat view, open the tools picker (Agent mode) and check that the `ops` tools appear. ## Test it In the Chat view (Agent mode), try: ``` Use ops to list available slices ``` You should see a `list_slices` tool call and get back your team's slices. ## Troubleshooting | Problem | Solution | | ----------------------------- | -------------------------------------------------------------------------- | | `ops` not in tools picker | Run **MCP: List Servers**, start `ops`, and ensure you're in Agent mode | | Authorization didn't complete | Restart `ops` from **MCP: List Servers** and finish login + team selection | | Wrong team's data | Remove `ops` from `mcp.json`, re-add it, and select a different team | | Server won't start | Confirm the URL and that `type` is `http` in `mcp.json` | ## Switching teams Each connection is scoped to one team. To switch, remove and re-add the `ops` server, then select the new team during sign-in. # Call Types Source: https://docs.oneperfectslice.ai/product/call-types OnePerfectSlice automatically classifies every call by type using AI. Classification drives everything downstream — which summary template is used, which scorecard criteria apply, and how the call is grouped for analysis. ## How it works Classification is fully AI-powered. No rules to maintain, no manual tagging. 1. The transcript is analyzed by AI alongside your team's configured call types 2. A **primary type** is assigned (required), plus an optional **secondary type** 3. An **AI-generated title** is created for the call 4. A **rationale** explains why the call was classified that way Once classified, the matched summary and scorecard templates are automatically applied — no manual step required. ## System call types OnePerfectSlice includes 11 built-in call types, each with its own summary and scorecard template: | Call Type | What the summary surfaces | | ------------------------------ | ------------------------------------------------------------------------------------------------------------ | | **Discovery Call** | Pain points, decision criteria, budget signals, timeline, competitive landscape, buying team, and next steps | | **Demo Call** | Feature reactions, pain-to-feature mapping, objections addressed, concerns, and next steps | | **Customer Growth Call** | Expansion signals, growth obstacles, stakeholder dynamics, competitive signals, and enablement needs | | **Customer Follow-Up Call** | Momentum, remaining obstacles, stakeholder dynamics, and follow-up actions | | **Renewal Call** | Renewal signals, churn risks, value anchoring, competitive signals, and enablement needs | | **Onboarding & Training Call** | Training objectives, topics covered, key learnings, engagement level, and follow-up items | | **Support Call** | Support risks, feature gaps, workflow fit, resolution signals, and enablement needs | | **Internal Strategy Call** | High-level business planning and positioning | | **Internal Execution Call** | Day-to-day operations, process improvements, and tactical planning | | **Internal Coordination Call** | Post-engagement alignment and internal action items | | **Miscellaneous Call** | General-purpose summary for calls that don't fit other categories | Teams can also create **custom call types** with their own summary and scorecard templates tailored to their specific workflow. ## Why it matters Without classification, every call gets the same generic treatment. With it: * **Summaries are tailored** — a discovery summary surfaces different context than a demo summary * **Scorecards are relevant** — reps are evaluated against criteria that match the type of conversation they're having * **Patterns emerge** — compare discovery calls to discovery calls, not discovery calls to proposal calls # Example: Demo Call Scorecard Source: https://docs.oneperfectslice.ai/product/examples/demo-scorecard A real example of the structured scorecard OnePerfectSlice generates for a demo call — evaluating pain reconnection, solution fit, objection handling, and next steps with G/Y/R ratings and coaching. This is an example scorecard generated from a demo call between a solutions engineer at **Vetro** (a B2B data integration platform) and the technical team at **Ridgeline Financial**, a Series C fintech company. All names, companies, and details are fictional. The structure and depth of the output reflect what OnePerfectSlice actually produces. *** ## BLUF (Bottom Line Up Front) The SE delivered a strong technical demo that connected well to Ridgeline's schema governance pain, but missed an opportunity to reconfirm the full pain landscape before jumping into features. The schema drift detection moment was a genuine turning point — David's production incident reference shifted the room from evaluation mode to envisioning mode. Two open items threaten momentum: the streaming throughput gap needs a benchmark within two weeks, and the PCI-DSS question was deferred without a clear owner or timeline. The deal has technical credibility but the clock is ticking against their Q3 budget lock. *** ## Next Moves * **Value Move**: Provision the benchmark environment this week and send David a preview of the test plan — don't wait for him to ask. Show throughput results at 2M+ transactions/day within two weeks to close the performance objection before it becomes a blocker. * **Objection Move**: Get the PCI-DSS mapping question to Vetro's security team today and schedule a CISO-to-CISO call within the next week. David flagged this as a blocker — every day without a scheduled security conversation is a day closer to the budget lock. * **Stakeholder Move**: Send David the build-vs-buy cost comparison he needs for the CFO conversation. He told you what he needs — deliver it before he has to ask again, and frame it around total cost of pipeline maintenance, not just licensing. *** ## Scorecard | Score | Element | Rationale | | ----- | ----------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | 🟡 | **Pain Reconnection** | The SE moved quickly into the demo without explicitly reconfirming the pain points from discovery. David's schema drift comment ("we had a production incident last month because of exactly this") surfaced pain organically, but the SE didn't build on it — didn't ask what the incident cost them, how long it took to resolve, or whether it affected downstream reporting. | | 🟢 | **Solution Fit Validated** | The schema drift detection demo was highly effective — it demonstrated the exact problem David described, using a scenario that mapped to a real incident. The lineage visualization engaged a previously silent stakeholder. The SE made a strong connection between Vetro's capabilities and Ridgeline's specific technical environment (Kafka, Snowflake, PostgreSQL). | | 🟡 | **Use Cases Confirmed** | The primary use case (replace custom Kafka-to-Snowflake pipeline) was validated through the demo, and schema governance was confirmed as a strong fit. However, the SE didn't confirm whether there are secondary use cases beyond the pipeline replacement. The compliance use case emerged organically but wasn't formally scoped or confirmed as a requirement. | | 🟡 | **Objections Addressed Live** | The SE handled the throughput objection well tactically — acknowledged the gap, proposed a benchmark, and David accepted. But the PCI-DSS question was deferred without a clear next step: no owner assigned, no timeline committed, and no interim answer provided. David explicitly flagged this as a blocker for the CISO, and the SE left it hanging. | | 🔴 | **Next Steps** | The demo ended without clearly defined next steps. The benchmark was agreed upon conceptually but no timeline was set. The PCI-DSS security review was identified as needed but no call was scheduled. The build-vs-buy analysis David mentioned wanting was acknowledged but not committed to. Three open items, zero with dates or owners attached. | *** Learn how OnePerfectSlice evaluates every call element-by-element with evidence-backed scoring and coaching suggestions. # Example: Demo Call Summary Source: https://docs.oneperfectslice.ai/product/examples/demo-summary A real example of the structured summary OnePerfectSlice generates for a demo call — showing pain themes, value hooks, and GTM-ready slices including demo moments, feature reactions, and objections. This is an example summary generated from a demo call between a solutions engineer at **Vetro** (a B2B data integration platform) and the technical team at **Ridgeline Financial**, a Series C fintech company. All names, companies, and details are fictional. The structure and depth of the output reflect what OnePerfectSlice actually produces. *** ## BLUF (Bottom Line Up Front) Ridgeline's Head of Data is evaluating Vetro to replace their custom Kafka-to-Snowflake pipeline — the schema drift detection and compliance lineage features landed strongly, but a latency gap on streaming ingestion and an unresolved PCI-DSS mapping question need to be closed before their Q3 budget locks in six weeks. *** ## Pain Themes * **Schema drift across services**: No centralized schema governance — a breaking change in a downstream service caused a production incident last month that took two days to diagnose. * **Manual compliance documentation**: Their compliance analyst currently builds lineage reports manually in spreadsheets for quarterly audits, taking roughly a week each cycle. * **Pipeline fragility**: Custom Kafka consumers are maintained by a single engineer — no redundancy, no monitoring, and any changes require a full deploy cycle. *** ## Value Hooks * **Schema drift detection**: When a breaking change was introduced in the live demo, the system flagged and blocked it automatically — David said "we had a production incident last month because of exactly this." * **Automated lineage visualization**: The compliance analyst on the call unmuted for the first time to ask detailed questions about export formats and GRC integration — strong engagement signal. * **Managed connector library**: The prospect is attracted to reducing maintenance burden on their custom pipeline, though they need throughput benchmarks before committing. *** ## Additional Context | Element | Value | | ------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | **Demo Moment That Landed** | Schema drift detection demo — when the system automatically blocked a breaking change, David referenced a real production incident it would have prevented. The room shifted from evaluating to envisioning. | | **Features Highlighted** | Schema Registry with Drift Detection (strongest reaction), Automated Lineage Visualization (engaged compliance stakeholder), Managed Connectors for Kafka/Snowflake/PostgreSQL (positive but throughput concerns), Role-Based Access Controls (brief mention, no strong reaction). | | **Integrations** | Kafka (primary ingestion), Snowflake (warehouse), PostgreSQL (operational DB). Need to validate throughput at 2M+ transactions/day for Kafka connector specifically. | | **Competitive Landscape** | Build vs. buy decision — current custom pipeline is the primary "competitor." No other vendor evaluation in progress. | | **Persona Insight** | Head of Data (David Park) is a technical buyer who makes decisions based on benchmark data, not demos. He'll champion internally but needs hard numbers to present to the CFO for a build-vs-buy analysis. | | **Stakeholder Insight** | Unnamed compliance analyst has quiet influence on security requirements. CISO is a gatekeeper (not on call) who will need a direct security-to-security conversation. CFO wants a build-vs-buy cost analysis. | | **Objection or Concern Raised** | Streaming throughput: live demo showed \~3 second latency vs. their current sub-second at 2M+ transactions/day. David was direct — "we can't regress on performance." PCI-DSS mapping question deferred to security team. | | **Asset Need** | One-page build-vs-buy cost comparison focused on total cost of ownership for custom Kafka pipeline maintenance vs. Vetro managed connectors. | *** Learn how OnePerfectSlice generates structured summaries automatically for every call, tailored by call type. # Example: Discovery Call Scorecard Source: https://docs.oneperfectslice.ai/product/examples/discovery-scorecard A real example of the structured scorecard OnePerfectSlice generates for a discovery call — showing element-by-element evaluation with G/Y/R confidence ratings, rationale, and coaching suggestions. This is an example scorecard generated from a discovery call between a sales rep at **Vetro** (a B2B data integration platform) and a prospect at **Lumen Health**, a mid-market healthcare analytics company. All names, companies, and details are fictional. The structure and depth of the output reflect what OnePerfectSlice actually produces. *** ## BLUF (Bottom Line Up Front) Strong discovery with clear pain and urgency — the rep surfaced real operational cost and a board-level deadline, but left the decision process underexplored. Rachel is a solid champion candidate, though the CISO and CFO are unknown variables. The deal has momentum, but without a mapped decision process and timeline to close, it risks stalling after the technical deep-dive. Next two weeks are critical. *** ## Next Moves * **Decision Move**: Schedule a call with Rachel to map the full approval chain — who signs, what procurement steps exist, and whether the CISO review can run in parallel with the technical evaluation. Target: this week. * **Value Move**: Build a cost-of-inaction model showing the \$180K+ annual reconciliation overhead plus compliance risk exposure — Rachel needs hard numbers for the CISO and CFO conversations. * **Champion Move**: Send Rachel a one-page internal business case template she can customize — she's willing to champion but needs ammunition that speaks to the board's language, not engineering language. *** ## Scorecard | Score | Element | Rationale | | ----- | -------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | 🟢 | **Situation** | Rep established a thorough picture of Lumen's current state — three ETL tools, two-state expansion, and the compliance gap. Rachel volunteered specifics (tool names, team structure, analyst headcount) without heavy prompting, suggesting the rep created a safe space for disclosure early. | | 🟢 | **Pain** | Three distinct pain themes surfaced with concrete business impact: fragmented pipelines causing data conflicts, compliance exposure from the state expansion, and 30% analyst time burned on reconciliation. The rep connected each pain to operational cost, not just inconvenience. Rachel's "keeps me up at night" quote about audit lineage confirms this is felt pain, not theoretical. | | 🟡 | **Impact** | The rep quantified reconciliation overhead (\$180K+ in analyst time) and surfaced the board-level accountability dimension, but didn't push deeper on the compliance exposure — what's the actual cost or consequence if they fail the Q4 audit? The impact of inaction on the state expansion was left vague. | | 🟢 | **Critical Event** | Clear and credible — the October board meeting is a hard deadline, the Q4 reporting season creates a fixed window, and the two-state expansion is already in motion. The rep identified the board as the forcing function and confirmed Rachel is personally accountable for showing progress. | | 🔴 | **Decision Process** | The rep identified three stakeholders (Rachel, CISO, CFO) but didn't map how they interact — who signs, what order, whether there's a procurement process, or how long security reviews typically take. The CISO and CFO are names without context. No discussion of competing priorities for budget or internal politics that could slow approval. | *** Learn how OnePerfectSlice evaluates every call element-by-element with evidence-backed scoring and coaching suggestions. # Example: Discovery Call Summary Source: https://docs.oneperfectslice.ai/product/examples/discovery-summary A real example of the structured summary OnePerfectSlice generates for a discovery call — showing pain themes, value hooks, and GTM-ready slices automatically extracted from the conversation. This is an example summary generated from a discovery call between a sales rep at **Vetro** (a B2B data integration platform) and a prospect at **Lumen Health**, a mid-market healthcare analytics company. All names, companies, and details are fictional. The structure and depth of the output reflect what OnePerfectSlice actually produces. *** ## BLUF (Bottom Line Up Front) Lumen Health is running three separate ETL tools across clinical, claims, and operational data and needs a unified pipeline with compliance-ready lineage before Q4 reporting season — driven by board pressure and two new state expansions that trigger audit requirements they can't currently meet. *** ## Pain Themes * **Fragmented pipeline ownership**: Three separate ETL tools managed by different teams with no shared schema registry, leading to conflicting data definitions across departments. * **Compliance exposure at scale**: Expanding into two new states triggers new reporting requirements, and the current stack can't produce audit-ready lineage documentation. * **Manual reconciliation overhead**: Two full-time analysts spend roughly 30% of their time reconciling mismatches between clinical and claims data before it reaches the analytics layer. *** ## Value Hooks * **Unified schema governance**: Vetro's shared schema registry directly addresses the conflicting data definitions problem — Rachel referenced this as "the thing that would save us the most time." * **Healthcare-specific connectors**: Pre-built connectors for HL7, FHIR, and state Medicaid systems would eliminate the 6-8 week onboarding cycle for new data sources. * **Audit-ready lineage**: Automated lineage documentation maps directly to the SOC 2 and state compliance requirements they're worried about for Q4. *** ## Additional Context | Element | Value | | ------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | **Messaging Hook** | "We lose two weeks every quarter just reconciling data before anyone can analyze it" — compliance pressure plus operational waste creates dual urgency. | | **Repositioning Angle** | Frame data integration as a compliance tool, not just an engineering convenience — the audit lineage angle resonated more than the pipeline consolidation pitch. | | **Persona Insight** | VP of Engineering (Rachel Mendez) cares most about reducing team overhead and board-level accountability — she needs to show progress on unification before the October board meeting. | | **Voice of Customer** | "The thing that keeps me up at night isn't the pipelines breaking — it's that we can't prove to auditors where the data came from." | | **Asset Need** | Healthcare-specific compliance guide showing how lineage documentation maps to SOC 2 and state Medicaid audit requirements. | | **Competitive Landscape** | Currently trialing Airbyte (open-source, cost appeal but gaps in healthcare connectors). Previously evaluated Fivetran (rejected on price and connector coverage). | | **Integrations** | HL7 feeds from two hospital networks, FHIR API from EHR vendor, three state Medicaid portals, Snowflake as destination warehouse. | *** Learn how OnePerfectSlice generates structured summaries automatically for every call, tailored by call type. # Example: Renewal Call Scorecard Source: https://docs.oneperfectslice.ai/product/examples/renewal-scorecard A real example of the structured scorecard OnePerfectSlice generates for a renewal call — evaluating adoption review, objection handling, risk identification, and decision process with G/Y/R ratings and coaching. This is an example scorecard generated from a renewal call between an account manager at **Vetro** (a B2B data integration platform) and the data team at **Clarion Logistics**, an existing customer approaching their annual renewal. All names, companies, and details are fictional. The structure and depth of the output reflect what OnePerfectSlice actually produces. *** ## BLUF (Bottom Line Up Front) The AM navigated a tricky renewal conversation — strong on demonstrating adoption and value, but too passive on the Salesforce connector objection and too slow to convert Marcus's expansion interest into a concrete proposal. The deal will likely close, but the AM left money on the table by not bundling the streaming tier into the renewal conversation, and let Marcus control the pricing narrative instead of reframing around value delivered. The CTO's "bring back options" comment is a negotiation tactic, not a churn signal — but only if the AM moves fast on the restructured proposal. *** ## Next Moves * **Decision Move**: Confirm the renewal approval chain with Marcus this week — does the CTO need to sign off formally, or is Marcus's signature sufficient? If the CTO is involved, offer to join that conversation directly rather than letting Marcus relay the pricing proposal secondhand. * **Value Move**: Build the restructured pricing proposal around value delivered, not volume consumed — lead with the 3-to-7 pipeline growth, the business-critical ops workflows, and the projected ROI of the streaming tier. Don't let this become a price negotiation disconnected from impact. * **Champion Move**: Give Marcus the proposal in a format he can present as his own win — he needs to show the CTO that he negotiated a better deal and secured a fix for the connector issues. Make him look good internally and he'll close this himself. *** ## Scorecard | Score | Element | Rationale | | ----- | ----------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | 🟢 | **Adoption Reviewed** | The AM effectively surfaced strong adoption metrics — growth from 3 to 7 pipelines, business-critical daily ops workflows, and organic expansion without sales involvement. Marcus's "revolt" comment was a clear signal of deep dependency. The AM used these data points to anchor the renewal conversation in value delivered rather than just contract terms. | | 🟡 | **Objections Addressed Live** | The AM acknowledged the Salesforce connector issues and referenced the recent engineering fix, which was the right move. However, the response was reactive — the AM waited for Marcus to raise the issue rather than proactively addressing it upfront. The dedicated Slack channel offer was a good tactical move but felt improvised rather than prepared. On pricing, the AM accepted Marcus's framing ("reset the economics") without pushing back on the value already delivered. | | 🟡 | **Deal Risks Identified** | The AM picked up on the CTO "bring back options" signal and the pricing sensitivity, but didn't probe deeper on either. Is the CTO genuinely evaluating alternatives, or is this a negotiation tactic? The AM also didn't address what happens if the connector issues recur during the renewal window — that's the highest-probability risk and it wasn't discussed. | | 🟡 | **Decision Process** | The AM established that Marcus has renewal authority and the CTO has oversight, but didn't map the full process — is there a procurement step? Does finance need to approve a pricing restructure? How long does the CTO typically take to review proposals? The 45-day window is known, but the AM didn't work backward from it to build a timeline with milestones. | | 🔴 | **Next Steps** | The call ended with a single vague commitment — the AM will send a restructured proposal "within one week." No specific date, no format confirmed, no follow-up meeting scheduled. The expansion conversation (streaming tier) was acknowledged but not converted into a next step. Three opportunities, zero locked on the calendar. | *** Learn how OnePerfectSlice evaluates every call element-by-element with evidence-backed scoring and coaching suggestions. # Example: Renewal Call Summary Source: https://docs.oneperfectslice.ai/product/examples/renewal-summary A real example of the structured summary OnePerfectSlice generates for a renewal call — showing renewal signals, renewal risks, decision dynamics, and deal-level slices. This is an example summary generated from a renewal call between an account manager at **Vetro** (a B2B data integration platform) and the data team at **Clarion Logistics**, an existing customer approaching their annual renewal. All names, companies, and details are fictional. The structure and depth of the output reflect what OnePerfectSlice actually produces. *** ## BLUF (Bottom Line Up Front) Clarion Logistics intends to renew but is leveraging three quarters of Salesforce connector issues to negotiate a pricing restructure — strong platform dependency and organic adoption growth from 3 to 7 pipelines signal retention, while the CTO's request to "bring back options" is a soft competitive signal that adds urgency to closing the restructured deal before the 45-day expiration window. *** ## Renewal Signals * **Deep platform adoption**: Expanded from 3 to 7 active pipelines organically — no sales push required. Operations team built daily warehouse replenishment reporting on top of Vetro that is now business-critical. * **Team dependency**: Marcus said "our ops team would revolt if we pulled Vetro out at this point" — the platform is embedded in their daily workflows with no viable alternative in place. * **Unprompted expansion interest**: Marcus asked about the real-time streaming tier without prompting — they want to track inventory movement across 12 distribution centers, currently running batch jobs every 4 hours. *** ## Renewal Risks * **Salesforce connector frustration**: Three support tickets in the past quarter related to sync failures and field mapping issues on custom objects — Marcus described this as "the one thing that keeps coming up" in their internal platform reviews. * **CTO pressure to evaluate options**: Marcus mentioned the CTO asked him to "bring back options" — no formal evaluation underway, but the door is open if Vetro doesn't address their concerns during the renewal window. * **Pricing sensitivity**: Marcus framed the renewal as an opportunity to "reset the economics" — they feel per-pipeline pricing doesn't reflect the volume they're running through the platform. *** ## Decision Dynamics * **Decision authority**: Marcus has authority to approve the renewal but indicated the CTO is watching the outcome, particularly on pricing and the Salesforce connector resolution. * **Pricing structure preference**: Marcus wants to move from per-pipeline to volume-based pricing — needs a restructured proposal he can present to the CTO as a win within one week. * **Timeline constraint**: Contract expires in 45 days. Marcus wants the restructured proposal within one week, CTO conversation likely within two weeks after that. *** ## Additional Context | Element | Value | | -------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | **Proposal Pricing** | Current per-pipeline model is the main negotiation point. Marcus wants volume-based pricing that accounts for their growth from 3 to 7 pipelines. Streaming tier would be incremental budget. | | **Budget** | No specific dollar amount discussed. Budget exists for renewal — the negotiation is about structure (per-pipeline vs. volume) rather than total spend. Streaming tier expansion would require separate approval. | | **Alternatives Discussed** | No specific competitors named. CTO's "bring back options" is the only competitive signal — more of a negotiation lever than an active evaluation. | | **Decision Process** | Marcus approves renewal, CTO has oversight. Two-step process: (1) agree on restructured pricing, (2) CTO sign-off on the package including connector resolution commitment. | | **Deal Risks** | If Salesforce connector issues recur during the renewal window, it undermines the reliability narrative. Pricing negotiation could stall if the restructured proposal doesn't land within the one-week window Marcus requested. | | **Timeline Urgency** | Contract expires in 45 days. Marcus wants restructured proposal within one week. CTO conversation expected within two weeks. Need verbal commitment before the 30-day expiration threshold. | *** Learn how OnePerfectSlice generates structured summaries automatically for every call, tailored by call type. # Scorecards Source: https://docs.oneperfectslice.ai/product/scorecards OnePerfectSlice automatically scores every call against a structured evaluation template — matched to the call type. Scorecards don't just tell you how a call went; they produce consistent, comparable context across every rep and every conversation — ready for your team, agents, and workflows. ## How scorecards work 1. A call is ingested from your [connected recording platform](/integrations/index) or uploaded directly 2. The call is [automatically classified](/product/call-types) into a type (discovery, demo, etc.) 3. The scorecard template matched to that call type is applied 4. AI evaluates the transcript element by element, scoring each one with evidence from the conversation 5. A structured scorecard is generated and available to your team, agents, and workflows This happens automatically for every call — no manual trigger required. Each element is scored on a three-point scale: | Score | What it means | | -------------- | --------------------------------------------- | | **G** (Green) | Rep covered this element thoroughly | | **Y** (Yellow) | Partially addressed — key aspects were missed | | **R** (Red) | Not addressed | For each score, the AI provides: * **Rationale** — why it scored that way, with direct quotes from the call * **Coaching suggestion** — specific, actionable advice for improvement ## Structured by call type Every call type has its own scorecard template — so a discovery call is evaluated against different elements than a demo or a renewal call. Because every call of the same type is scored against the same criteria, you can compare across reps, track improvement over time, and identify team-wide patterns — and your agents and workflows can consume the same structured context programmatically. For example: * **[Discovery Call scorecards](/product/examples/discovery-scorecard)** evaluate decision criteria explored, pain identification, and initiative maturity * **[Demo Call scorecards](/product/examples/demo-scorecard)** evaluate whether the demo mapped to pain, objections addressed live, and next steps * **[Renewal Call scorecards](/product/examples/renewal-scorecard)** evaluate value anchoring, risk handling, and competitive awareness See [Call Classification](/product/call-types) for the full list of call types and what each template covers. ## Access this context anywhere Scorecards are available in the app, through the [API](/api/introduction), and to AI agents via the [MCP server](/mcp/overview). Browse, filter, and drill into element-by-element scores. Search and retrieve scorecards programmatically. Ask Claude to find scorecards — "Show me scorecards where reps scored Red on objection handling." ## Frequently asked questions **How does the scoring work?** Each element is scored on a three-point scale: Green (rep covered this element thoroughly), Yellow (partially addressed — key aspects were missed), or Red (not addressed). For each score, the AI provides a rationale with direct evidence from the call and a coaching suggestion for improvement. Green-rated elements receive no coaching suggestion. See [example scorecards](/product/examples/discovery-scorecard) for the full output. **What elements are evaluated on a scorecard?** It depends on the call type. Discovery call scorecards evaluate Situation, Pain, Impact, Critical Event, and Decision Process. Demo call scorecards evaluate Pain Reconnection, Solution Fit, Use Cases Confirmed, Objections Addressed Live, and Next Steps. Renewal call scorecards evaluate Adoption Reviewed, Objections Addressed, Deal Risks Identified, Decision Process, and Next Steps. Each call type has its own evaluation criteria. **Are scorecards generated automatically?** Yes. Every call that flows into OnePerfectSlice is automatically classified by type and scored against the matched scorecard template. No manual trigger or manager review required. **How are scorecards different from summaries?** Summaries extract structured context from a call — what was discussed, what matters, and what to do next. Scorecards evaluate how well the rep executed the call — element by element, with evidence-backed ratings and coaching suggestions. Both are generated automatically for every call. **Can I use scorecards to track team-wide patterns?** Yes. Because every call of the same type is scored against the same criteria, you can compare across reps and track improvement over time. If multiple reps are scoring Red on Decision Process, that's a team-wide coaching opportunity — not a one-off gap. **Can I access scorecards programmatically?** Yes. Scorecards are available through the [REST API](/api/endpoints#posts) and the [MCP server](/mcp/tools#search_posts). You can filter by call type, date range, rep, company, and CRM fields. # Slices Source: https://docs.oneperfectslice.ai/product/slices Slices are AI analysis templates that surface structured, evidence-backed context tied to a specific business goal and role. OnePerfectSlice includes 27 pre-built slices across pipeline, deals, expansion, and product. Where [summaries](/product/summaries) and [scorecards](/product/scorecards) focus on a single call, slices connect the dots across many conversations to surface trends, gaps, and opportunities. ## How to run a slice View available slices, filtered by business objective or role. Narrow the scope: date range, call types, reps, companies, CRM deals or stages. Check how many calls match before committing — previews are free. Start the analysis. Runs execute in the background — you can leave and come back. View structured output with a quality score. Every insight is backed by direct quotes from your calls with source attribution and confidence scores. ## Slice catalog OnePerfectSlice includes 27 pre-built slices organized by business objective. ### Grow Pipeline Slices for demand gen, content marketing, and go-to-market teams. | Slice | Description | Best for | | ------------------------- | ----------------------------------------------------------------------------------------------------------- | ----------------------------- | | **Personas Identified** | Identify buyer personas based on roles, responsibilities, and motivations from conversations | Product Marketing, Demand Gen | | **Segments Observed** | Identify company segments based on firmographics and buyer personas from conversations | Product Marketing, Demand Gen | | **Category Entry Points** | Identify the triggers and events that cause prospects to enter the market | Demand Gen, Marketing Ops | | **Positioning Angle** | Analyze positioning based on competitive landscape and unique selling points | Product Marketing | | **Messaging Hooks** | Generate punchy ad headlines and messaging hooks by combining customer pain points with hook type selection | Demand Gen, Content Marketing | | **Campaign Ideas** | Design campaign concepts by combining customer evidence with creative strategy selection | Demand Gen, Marketing Ops | | **Content Ideas** | Generate specific, shareable content pieces calibrated to buying stage and market position | Content Marketing, Demand Gen | ### Win Deals Slices for sales enablement, sales ops, and frontline managers. | Slice | Description | Best for | | -------------------------- | ------------------------------------------------------------------------------------------------- | ----------------------------------- | | **Alternatives Evaluated** | What buyers compare you against — competitors, DIY, status quo, and do-nothing options | Sales Enablement, Product Marketing | | **Decision Criteria** | What factors buyers weigh when choosing — their priorities, evaluation methods, and your position | Sales Enablement, Sales Ops | | **Objection Patterns** | Common objections, how they surface, and proven responses | Sales Enablement | | **Pricing Dynamics** | Price sensitivity patterns, discount triggers, and value justification approaches that work | Sales Ops, Revenue Leadership | | **What's Working** | Successful rep behaviors and techniques to replicate across the team | Sales Enablement, Sales Managers | | **What's Breaking** | Rep behaviors that are failing or causing deals to stall | Sales Enablement, Sales Managers | | **Talk Tracks That Win** | Specific phrases and approaches that close deals | Sales Enablement | | **Enablement Priorities** | Prioritized enablement actions to improve win rates and sales effectiveness | Sales Enablement, Sales Ops | | **Asset & Content Gaps** | Missing sales assets or content that would improve close rates | Sales Enablement, Content Marketing | | **Skills to Develop** | Prioritized skill gaps to address through coaching or training | Sales Enablement, Sales Managers | ### Expand Accounts Slices for account management, customer success, and expansion teams. | Slice | Description | Best for | | ------------------------ | ----------------------------------------------------------------- | ------------------------------------- | | **Expansion Signals** | Observable indicators that customers are ready for expansion | Account Management, Customer Success | | **Stakeholder Mapping** | Key stakeholders who can influence or approve expansion decisions | Account Management | | **Untapped Use Cases** | Use cases where customers could benefit but haven't adopted yet | Account Management, Product Marketing | | **Adoption Barriers** | Obstacles preventing deeper adoption or expansion | Customer Success, Product | | **Expansion Priorities** | Ranked expansion opportunities for account managers to pursue | Account Management, Sales Ops | | **Expansion Messaging** | Messages and value props that resonate for upsell conversations | Account Management, Product Marketing | ### Improve Product Slices for product management, product marketing, and engineering leadership. | Slice | Description | Best for | | ---------------------------- | --------------------------------------------------------------------------------------------- | ------------------------------------- | | **Feature Feedback** | Customer reactions to existing features — what resonates vs. needs attention | Product Management | | **Competitive Intelligence** | Competitor mentions, feature comparisons, and win/loss insights | Product Marketing, Product Management | | **Roadmap Priorities** | Prioritized feature requests and strategic improvements for the product roadmap | Product Management, Engineering | | **Quick Wins** | Small improvements with disproportionate customer impact requiring minimal engineering effort | Product Management, Engineering | ## Access slices Slices are available programmatically through the [API](/api/introduction), by agents through the [MCP server](/mcp/overview), and through the slices page within the app. List slices, preview counts, start runs, and retrieve results and evidence. Ask Claude to run a slice — it handles the full workflow in one tool call. # Summaries Source: https://docs.oneperfectslice.ai/product/summaries OnePerfectSlice automatically generates a structured summary for every call — tailored by call type so discovery calls, demos, and QBRs each surface the context that matters most. ## How summaries work 1. A call is ingested from your [connected recording platform](/integrations/index) or uploaded directly 2. The call is [automatically classified](/product/call-types) into a type (discovery, demo, QBR, etc.) 3. The summary template matched to that call type is applied 4. AI processes the transcript alongside any available CRM context 5. A structured summary is generated and available to your team, agents, and workflows This happens automatically for every call — no manual trigger required. ## Structured by call type This is the key: **every call type has its own summary template.** A discovery call summary surfaces different context than a demo summary or a QBR summary. Because every call of the same type follows the same structure, you can compare consistently across reps, accounts, and time periods. That's what makes the context useful at scale — for your team, your workflows, and your AI agents. For example: * **[Discovery Call summaries](/product/examples/discovery-summary)** surface pain points, decision criteria, budget signals, and competitive landscape * **[Demo Call summaries](/product/examples/demo-summary)** capture feature reactions, objections addressed, and next steps * **[Renewal Call summaries](/product/examples/renewal-summary)** focus on renewal signals, churn risks, and value anchoring See [Call Classification](/product/call-types) for the full list of call types and what each template covers. ## Access this context anywhere Summaries are available in the app, through the [API](/api/introduction), and to AI agents via the [MCP server](/mcp/overview). Browse, filter, and search summaries from the posts feed. Search and retrieve summaries programmatically. Pipe them into Zapier, n8n, or your own tools. Ask Claude to find summaries — "Show me summaries from Jane's discovery calls this week." ## Frequently asked questions **What does a summary include?** Every summary starts with a BLUF (Bottom Line Up Front) — a one-sentence overview of what matters most. Below that, structured sections are tailored to the call type. Discovery summaries include pain themes, value hooks, and GTM-ready slices like messaging hooks, competitive landscape, and voice-of-customer quotes. Demo summaries capture feature reactions, objections, and demo moments. Renewal summaries surface renewal signals, churn risks, and decision dynamics. See [example summaries](/product/examples/discovery-summary) for the full output. **Are summaries generated automatically?** Yes. Every call that flows into OnePerfectSlice — whether from Gong, Fireflies, Fathom, or uploaded directly — is automatically classified by type and summarized using the matched template. No manual trigger required. **Why do different call types get different summaries?** A discovery call surfaces different context than a demo or a renewal. By tailoring the template to the call type, each summary extracts the context that actually matters for that type of conversation. This also makes summaries comparable — every discovery summary follows the same structure, so you can compare across reps, accounts, and time periods. **Can I create custom summary templates?** Yes. Teams can create custom call types with their own summary templates tailored to specific workflows. See [Call Types](/product/call-types) for details. **How do I access summaries programmatically?** Summaries are available through the [REST API](/api/endpoints#posts) and the [MCP server](/mcp/tools#search_posts). You can filter by call type, date range, rep, company, and CRM fields. # Quickstart Source: https://docs.oneperfectslice.ai/quickstart This guide walks you through the first-time experience in OnePerfectSlice. By the end, you'll have calls flowing in, AI-generated summaries and scorecards appearing automatically, and will have run your first slices. ## Set up your workspace Sign up at [app.oneperfectslice.ai](https://app.oneperfectslice.ai) with your work email. You'll receive a verification email — click the link to confirm. On first login, you'll be prompted to enter your name, create your organization, and set up your first team. Teams are how OnePerfectSlice scopes data — each team sees only its own calls and insights. From **Org Settings**, invite colleagues by email. They'll receive a link to join your team directly — no separate org creation needed. Got an invite link? Click it and follow the prompts — you'll skip org creation and land directly in your team's workspace. ## Import your call recordings OnePerfectSlice needs access to your call transcripts. You can either connect a recording platform or upload calls directly. Navigate to **Org Settings → Integrations** in the app. Choose your call recording source and follow the authorization flow: | Platform | How to connect | | ------------- | ---------------------------------------------------- | | **Gong** | API key pair — paste your Gong access key and secret | | **Fireflies** | API key — paste your Fireflies API key | | **Fathom** | API key — paste your Fathom API key | Don't use one of these? You can also **upload calls directly** from the /transcripts page in the app. Once connected, OnePerfectSlice pulls in your existing calls and begins processing new ones as they happen. Each call is [automatically classified](/product/call-types) by type (discovery, demo, QBR, etc.). ## Review your posts Within minutes of connecting, AI-generated posts start appearing for your team. Every processed call produces: * **A summary** — topics, decisions, action items, and next steps * **A scorecard** — structured evaluation against your team's methodology Navigate to the **Posts** section in the app. You'll see a feed of summaries and scorecards, most recent first. Use filters to narrow by date, call type, rep, or company. Click into any post to see the full structured output. Every post is tied to a **call type** — and each call type has its own summary and scorecard template. This means every discovery call is summarized the same way, every demo is scored against the same criteria. That consistency is what enables comparison and pattern matching at scale. ## Run your first slice Slices go beyond individual calls — they analyze patterns across many conversations to answer a specific business question. Navigate to the **Slices** section. You'll see available analysis templates organized by business objective: pipeline, deals, expansion, and product. Choose one that matches what you're trying to learn. For example: * **Objection Patterns** — what's blocking deals from closing? * **Competitive Intelligence** — who's coming up in conversations and why? * **Campaign Ideas** — what campaigns should you run based on what customers are saying? Choose a date range and optionally filter by call type, rep, company, or CRM deal. Use **Preview Count** to see how many calls match before running. Click run. OnePerfectSlice analyzes the matched calls and returns structured insights backed by evidence — direct quotes from your conversations with confidence scores. ## What's next You're set up. From here: Learn about the 27 available analysis templates and how to get the most from them. Add HubSpot to enrich your analyses with deal and company context. Set up MCP so Claude can query your data conversationally.