Example applications for Corti products. Each project is self-contained with its own README and quick-start instructions.
Prerequisites: A Corti account and API credentials from Corti Console.
A pre-built Corti Web Component that can be embedded directly into any web application. The assistant manages the full clinical session experience — recording, transcription, fact review, and documentation — so the host application only needs to handle authentication and interaction lifecycle.
Ambient scribing lets the system listen passively to a clinical encounter and automatically produce structured documentation without requiring the clinician to dictate or type. Audio is streamed in real time, clinical facts are extracted as the conversation unfolds, and a document is generated at the end. Can support both in-person (single microphone with diarization) and virtual consultation (doctor and patient on separate audio channels).
Dictation gives clinicians direct control over the content, wording, and structure of their documentation through speech. Unlike ambient scribing — where the system listens in the background — dictation is intentional: the provider speaks directly into the record. Supports real-time transcription with interim results, spoken punctuation, and voice commands for navigation and editing. Complements ambient workflows for specialties or situations where direct control is preferred.
The Corti Agentic Framework is a modular AI system for building advanced agents that perform clinical and operational tasks. Agents are autonomous systems that can think, reason, and adapt — using AI to understand context, make decisions dynamically, and take actions. These examples demonstrate how to integrate Corti agents into custom applications using the A2A (Agent-to-Agent) protocol for real-time streaming communication, enabling chat interfaces and interactive assistants with the Corti SDK and modern frontend frameworks.
Corti APIs should never be called directly from the browser — doing so would expose client credentials. Instead, route requests through a server-side proxy that authenticates with Corti on behalf of the client. The proxy is also the right place to enforce your own access control, add rate limiting, or inject logging.
A reference showcase of the Corti SDKs (@corti/sdk · Corti.Sdk) covering every group of endpoints: auth flows, interactions, recordings, transcripts, facts, codes, templates, agents, documents, and WebSocket streaming. Not real applications — intended as a practical guide for developers integrating the SDK and as a validation tool when upgrading to a new SDK version.
| Example | Stack | Description |
|---|---|---|
| sdk/typescript/express-web-api/ | TypeScript, Express | REST API covering all SDK features |
| sdk/typescript/next-auth-examples/ | TypeScript, Next.js | Standalone demos of all four Corti OAuth flows (client credentials, ROPC, auth code, PKCE) |
| sdk/dotnet/web-api/ | C#, ASP.NET Core | Same scope as the Express example, in .NET |
| Example | Stack | Description |
|---|---|---|
| embedded-assistant/react/basic-example/ | TypeScript, React 19, Vite | React integration using @corti/embedded-react |
| embedded-assistant/vanilla-ts/basic-example/ | TypeScript | Web Component integration — no framework required |
| Example | Stack | Description |
|---|---|---|
| ambient-scribe/typescript/basic-example/ | TypeScript, Express | Single-mic or virtual consultation (doctor + patient channels); document generation from extracted facts |
| Example | Stack | Description |
|---|---|---|
| proxy/javascript/basic-example/ | JavaScript, Express | WebSocket proxy — browser connects to proxy, proxy authenticates with Corti SDK |
| Example | Stack | Description |
|---|---|---|
| dictation/typescript/basic-example/ | TypeScript, Express | Microphone streaming via the Corti SDK; interim + final transcripts; voice commands |
| dictation/typescript/web-component/ | TypeScript, Express | Four demos of @corti/dictation-web: basic, custom UI, styling, and token refresh |
| dictation/commands/ | Json | English voice-command phrase JSON (templates, dictation box, editing, lists, navigation, Roman numerals, selection) |
| Example | Stack | Description |
|---|---|---|
| agents/react/next-agent-chat/ | TypeScript, Next.js, React | AI agent chat interface with A2A streaming, AI SDK integration, and real-time messaging |