Features:
- Simple to self host (deploy to Vercel & Supabase)
- Use with any model, not just OpenAI
- Log agent runs and identify errors with full history
- Records completion requests, search and explore request history in the dashboard
- Latency and token analytics
Currently, a JS client is available with Python in the works.
- Doesn't add latency with an extra endpoint
- Langchain support
- Demo dashboard
- Search on agent and LLM pages
- Date filtering on agent and LLM pages
- Users Tracking
- Embeddings analytics
- Tools views
This project is licensed under the Apache 2.0 License.