LLM Tracing with MLflow AI Gateway
MLflow AI Gateway introduces LLM tracing capabilities for agentic applications, allowing developers to log LLM calls automatically. This feature aids in debugging by providing insights into the interactions between coding agents and LLMs. While it requires some code updates, it streamlines the process of monitoring and optimizing coding sessions.
- ▪MLflow AI Gateway allows automatic logging of LLM calls as traces.
- ▪Developers can instrument their agentic apps with MLflow for better debugging.
- ▪The gateway supports passthrough calling and is compatible with various SDKs.
Opening excerpt (first ~120 words) tap to expand
LLM tracing with MLflow AI GatewayApr 10, 2026#aiopsTracing. Wait, what?It's too easy to develop agentic apps, or use coding agents to help you with stuff. Claude Code, Gemini CLI, Codex, OpenCode, Copilot CLI and many more that I forgot to mention are popular for a reason. These coding agents would report how many tokens are consumed so you can plan your coding sessions as to not run out of usage credits.But do you ever wonder what's going on between the coding agent and LLM? Or when a user submits a query to your agent, which tools it used, and what's exactly being sent to the LLM? Agentic frameworks can and sometimes rewrite your original prompt, not to mention system prompts baked into the frameworks that you don't see unless you dive into the source code.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at Karnwong.
Discussion
0 commentsMore from Karnwong
