We're excited to announce a game-changing feature that's going to make your life easier when working with AI agents: Traces! 🎉 This powerful observability feature that lets you see exactly what your AI agents are doing under the hood. With zero setup required, you can monitor LLM calls, tool executions, and processing steps in real-time, making debugging and optimization easier in both dev and prod phases.
Ever wondered what's happening under the hood when your agent is processing a request? Now you can see it all in action, without writing a single line of monitoring code. That's right - total observability comes built-in with every agent you deploy on Blaxel!
Think of traces as a GPS for your agent's journey. Every time your agent processes a request, it creates a trace that shows you exactly what happened, when it happened, and how it all fits together. It's like having x-ray vision into your agent's thought process!
Each trace is made up of spans - think of them as the building blocks of your agent's journey. A span shows you:
Our traces come with some pretty neat features out of the box:
Zero setup required: Deploy from our console or SDK, and traces are automatically enabled
Complete visibility: See everything from LLM calls to tool executions in one clear view
Smart sampling: We automatically save traces for 10% of your executions to help you monitor performance
Debug mode: Need to trace a specific execution? Just add debug=true to your API call
