From Blind Spots to Merged PRs: Continuous Agentic Performance Optimization - May Walter, Hud

Performance issues silently pile up in mature codebases. Teams know things could be faster, but can never justify pausing feature work to investigate. You have to put engineers on it just to find out if there's something worth fixing, and the effort is completely unpredictable: it could take an hour or three weeks.

In this talk, we'll walk through a real case study of adding runtime intelligence to coding agents to enable continuous performance optimization in production. We'll cover the pain that led us here, the technical approach (agents analyzing real production context to surface high-ROI fixes scored by complexity and impact), and what we had to improve along the way to get reliable results.

This approach surfaced and fixed N+1 queries and missing database indexes within the first week, with measurable P90 latency improvements after deployment. Tech leads now receive actionable reports before sprint planning and can make decisions starting from the fix, not the problem.

If you're looking for a concrete, real-world example of integrating AI agents into the SDLC, this is it: not a demo, not a prototype, but merged PRs and better production performance.

Speakers:
- May Walter (Hud): May Walter is Co-Founder and CTO of Hud, where she builds the Runtime Code Sensor that gives coding agents real-time production intelligence, drawing on her background as a serial CTO, runtime internals expert, and adversarial cybersecurity researcher.
  LinkedIn: https://www.linkedin.com/in/may-walterr/