Introduction

The AfterTalk API gives you everything the web app does, over HTTP: upload a recording, get back a transcript with speaker roles, a summary, a category, action items, and template-driven analysis.

The mental model

Three things, in order:

1. Organization — owns your plan, your projects, and your API keys.
2. Project — a workspace for related calls. Categories, analysis templates, and speaker presets are configured per project, and every call belongs to one.
3. Call — one uploaded recording, plus everything the pipeline derived from it.

Almost every endpoint is scoped to a project:

/v1/projects/{projectId}/calls

How a call flows through the system

Uploading is asynchronous. You get a job id immediately, and the transcript appears when processing finishes:

POST /calls/upload  →  202 { jobId }
      ↓  transcription → PII redaction → classification → summary → template analysis
GET  /calls/jobs/{jobId}  →  { status: "COMPLETED", transcriptId }
GET  /calls/{transcriptId}  →  transcript, summary, category, action items

Poll the job until it reports a terminal status, or register a webhook and be told when it's done.

What you can do

• Upload calls and track processing
• Read transcripts, summaries, sentiment, categories, and action items
• Search across transcripts with full-text queries
• Manage projects, categories, analysis templates, and speaker presets
• Read dashboard metrics and cross-call insights
• Receive webhooks when a call finishes processing

Deleting projects and calls is deliberately not available over the API — do that in the app.

Next steps

Create a key in Settings → API keys, then follow the quickstart.