AI-native product development, inside a 35M-user platform.
In my last chapter at Ticketmaster I rebuilt how my organization developed product, with AI at the center, and got back into the code myself after spending most of my career leading from beside it. This is what that looked like.
I ran global mobile product for a platform with 35M+ monthly users. When AI-assisted development became real, I did not commission a task force. I went first: got the codebases on my machine, started shipping, and used the work to pull product, design, and engineering into a new way of operating.
AI Search, end to end
Prototyped a full AI-powered search experience for the app, from concept through working code. Not a spec handed to a team: a running prototype that let executives react to the real thing.
AI-powered UI and UX
Prototyped new AI-driven interface concepts that collapse funnel steps the industry has treated as fixed for a decade. The question stopped being how to optimize a step and became whether the step should exist.
The analytics rebuild
Re-architected app analytics with AI-assisted development, shipping tagging and instrumentation PRs myself across iOS and Android, the unglamorous work that was quietly taxing the team.
Performance instrumentation
Built a performance debug layer into the app: every home-screen module, every API call, its latency and payload. Found a module making three API calls where one would do, and shipped the fix at an engineering all-hands. Load time dropped to about 100 milliseconds.
The org shift
Roughly 35 PRs merged into the iOS and Android codebases, reviewed by senior engineers to the same bar as their own work. An AI upskilling program across product, design, and tech. And a set of honest, sometimes uncomfortable conversations about what product, design, and engineering each become next.
The value is not coding for its own sake. A working proof lets an executive team see the direction, test the assumptions, and make the consequential decision with evidence.