Watson.ai is a desktop application designed to record and summarize meetings, supporting English, French, and Chinese. It targets professionals who need to efficiently extract key information from their discussions. The primary benefit is automated summarization, saving users time and effort in reviewing meeting content.

How It Works

The application leverages external APIs for its core transcription and summarization functionalities. Users are required to provide API tokens for services like AssemblyAI and OpenAI API, with an optional integration for Affinity API. This approach allows Watson.ai to utilize state-of-the-art AI models for processing audio and generating summaries without needing to develop these capabilities in-house.

Quick Start & Requirements

• Install: Download the app from the provided link.
• Prerequisites:
  • MacOS or Windows operating system.
  • API tokens for transcription services (AssemblyAI, OpenAI API).
  • FFMPEG binary must be present and correctly linked in src-tauri/bin/ for the specific platform.
• Setup: Requires manual configuration of API tokens and FFMPEG binary path.
• Links: App Demo, Async process in Rust + communication with webview

Highlighted Details

• Supports meeting recording and summarization in English, French, and Chinese.
• Requires user-provided API tokens for transcription services.
• Bundles FFMPEG but requires manual path configuration.
• Developed initially as an internal tool, now being made public.

Maintenance & Community

The project is maintained by LatentDream, who is actively developing it and planning future releases. Community contributions for adding platform support are welcomed.

Licensing & Compatibility

The README does not explicitly state a license. The application is distributed as a downloadable binary, and its compatibility with closed-source linking is not specified.

Limitations & Caveats

The application is not signed, requiring users to bypass OS security settings. Recording meetings may have legal implications depending on jurisdiction, and users are solely responsible for compliance. The project is in active development, and API token management is a manual user responsibility.