Speech-to-Notion

Speech-to-Notion directly transcribes speech to a designated Notion page, offering a streamlined, real-time transcription experience. Built on the NestJS framework and featuring a robust frontend, this project supports various audio formats and languages.

Objective

Speech-to-Notion aims to provide a fast, user-friendly platform for transcribing audio directly to Notion pages. It uses the Gladia API for accurate transcription across multiple languages and formats.

Setup

Follow these steps to get started:

• Install Node.js and Yarn.
• Get a Gladia API key at https://app.gladia.io/home.
• Obtain a Notion integration API key by creating an integration at https://www.notion.so/my-integrations.
• Set up environment variables in a .env file at the project root, as detailed below.

Environment Variables

Configure the following in .env for smooth operation:

Backend

• GLADIA_API_KEY: Your Gladia API key for transcription.
• NOTION_API_KEY: Your Notion integration API key.

Frontend

• NEXT_PUBLIC_BACKEND_ENDPOINT=http://localhost:4000: Backend service URL.

Starting the Services

1. Clone the repo.
2. In the project root, run yarn install.
3. Start the frontend with yarn start:frontend.
4. Start the backend with yarn start:backend.

Both services must run together.

Frontend Service

The frontend allows users to interact with the transcription service. It's built with modern web technologies and communicates with the backend for audio processing.

Key Files

• src/index.js/ts: Entry point.
• src/hooks/useTranscriber.ts: Manages transcription.
• package.json: Scripts and dependencies.
• .gitignore: Files Git should ignore.

Backend Service

The backend manages transcription and real-time communication via WebSocket.

Key Files

• audio-transcription.gateway.ts: WebSocket and transcription management.

Conclusion

Speech-to-Notion is a powerful, user-friendly tool for real-time transcription to Notion. Set up the environment variables and start both services to begin transcribing.