Devtools, AI, Insights - Vinicius Dallacqua - performance.now() 2025

Everyone is shipping AI features or Agents, but few of those systems can reason with the complex reality of performance data. This session is a deep dive story into the journey of building a Performance AI assistant and agentic workflows on top of a fork of Chrome DevTools. We'll explore practical lessons from DevTools internals on how to communicate performance data to Agents: teaching LLMs how to interpret performance data; which signals matter most; and how to transform telemetry into generated actionable insights from an LLM. Building automated attribution on top of known performance metrics and data, and exploring how to leverage specialized agents with a few experiments using a custom MCP server so coding agents and perf data can collaborate for a better generative workflow. Slides: https://docs.google.com/presentation/d/1Hl1hUeuduMMGmVdcDjtxYQwu0s4_y2GOE3kaMLyi7c4/edit *** A session from performance.now(), the web performance conference. Each November in Amsterdam; see https://perfnow.nl for more information. Conference announcements in your inbox? https://perfnow.nl/subscribe