Spending an entire afternoon correlating Blackfire profiles, flame graphs, git code, and traffic data just to figure out why an endpoint is slow is a grind. Generic LLMs do not help much because they lack access to your actual code and deployment setup.
The Upsun Performance Agent automates this analysis:
- Runs as a task container directly inside your Upsun project environment
- Pulls Blackfire profile data, continuous profiling, traffic metrics, and git source code
- Returns a structured report with a root cause diagnosis and a suggested diff
Why it matters:
- You get actionable analysis instead of raw call graphs
- You bring your own API key to control model selection and cost
- Code and profiles stay inside your environment control
Check out the article to see how it works and how to set it up:
Top comments (0)