Automating Frontend Performance Optimization with AI Agents

As frontend applications grow larger and more complex, performance optimization becomes increasingly challenging. Developers often spend significant time identifying bottlenecks, analyzing metrics, reviewing code changes, and deciding which optimizations will have the biggest impact. In this talk, I'll share how I've been exploring AI agent workflows to make frontend performance optimization faster and more efficient. Instead of using AI only as a coding assistant, I use specialized AI workflows to analyze performance issues, investigate rendering problems, review bundle sizes, identify unnecessary re-renders, and suggest optimizations across React and Next.js applications. I'll walk through practical examples of how these workflows can assist during development, code reviews, and performance audits. We'll also discuss how AI agents can be integrated into existing engineering processes and CI/CD pipelines to help teams catch performance issues earlier and focus on higher-value work. The session will cover both the opportunities and limitations of AI-driven performance engineering, along with lessons learned from experimenting with these workflows in real-world frontend projects.