This project provides Omi, an open-source AI wearable designed to capture conversations, generate summaries, identify action items, and perform tasks. It targets users needing automatic, high-quality transcriptions for meetings, chats, and voice memos, offering a hands-free, intelligent audio processing solution.

How It Works

The Omi ecosystem comprises dedicated hardware (nRF chips for wearables, ESP32-S3 for Omi Glass), a Flutter-based mobile application, and a Python backend leveraging FastAPI, Firebase, Pinecone, Deepgram, OpenAI, and Silero VAD. AI personas are built using Next.js. The hardware captures audio, which is processed by the app for real-time transcription and can be sent to external services via webhooks. The backend handles advanced AI processing for summarization and action item extraction.

Quick Start & Requirements

• Primary install/run command:

  git clone https://github.com/BasedHardware/omi.git
  cd omi/app
  bash setup.sh ios # or android, macos
  

• Prerequisites: Mobile device (iOS/Android), webhook.site for testing custom apps.
• Links: Site, Download, Docs, Buy Dev Kit, Buy Glass Dev Kit.
• Estimated setup time: 2 minutes for the app, 1 minute for custom app creation.

Highlighted Details

• Open-source AI wearable hardware and software.
• Real-time transcription with webhook integration for custom workflows.
• Support for AI personas and SDKs in React Native, Swift, and Python.
• Developer-focused with options to build custom plugins and integrations.

Maintenance & Community

• Community: Active Discord server (http://discord.omi.me).
• Social: Follow on Twitter (https://x.com/kodjima33).
• Contributions: Contributions are encouraged via a dedicated guide, with paid bounties available for specific issues.

Licensing & Compatibility

• License: MIT License.
• Compatibility: The MIT license permits commercial use and integration into closed-source projects.

Limitations & Caveats

The project appears to be hardware-dependent, requiring specific Omi dev kits for full functionality. The setup script explicitly supports iOS, Android, and macOS, with no mention of Windows support.