This talk is a case study of how we built and evolved our AI-powered developer extension.

Our initial goal was simple: create a natural language interface that lets developers interact with our platform directly from their IDE. This would streamline onboarding for new application teams and reduce friction for platform engineers.

But as we rolled it out and gathered feedback, we realized its potential went far beyond convenience. By deeply integrating our internal knowledge—docs, forums, support tickets—we turned the tool into a guide, a troubleshooter, and even a code reviewer. We made it opinionated on purpose.

This talk will cover:
- How we identified our initial use case and then expanded it based on real-world usage.
- Why embedding strong opinions into AI tooling helps developers fall into the “pit of success” faster.
- How to build AI integrations that are genuinely useful, not just trendy.

Whether you're thinking about adding AI to your developer tools or refining what you already have, this session will offer practical lessons and hard-earned insights.