AI Native Engineering
How Meta teams are leveraging AI models to enhance productivity, quality, and understanding across the engineering organisation.
Meta is making a big, public bet on AI - and not just in our products. Teams across the company are building new tools to leverage best-in-class models to enhance productivity, quality, and understanding.
AI coding assistants and development tools are evolving rapidly, but the question that kept coming up in conversations with colleagues was: how do you actually integrate them into existing engineering workflows at scale? What works, what doesn’t, and what have we learned the hard way?
This talk explored our journey building and adopting AI-native engineering tools. I covered the practical patterns we’ve found effective for AI-assisted development, and how AI is starting to change code review, testing, and documentation in ways I didn’t initially expect. There’s a tension between embracing these tools and maintaining engineering excellence - I talked through how we’re thinking about that balance.
I also shared some of our false starts and lessons learned. Not everything we tried worked, and I find those stories are often more useful than the success stories when people are figuring out their own approach.
As always it was a pleasure to be part of a QCon event. Thank you if you came to listen or have watched the talk back online.