Context
Led the engineering organization behind Snapchat Discover & Stories during the platform's hyper-growth years, covering recommendation systems, creator and media workflows, story infrastructure, moderation systems, and content distribution for hundreds of millions of users.
What I built and led
- Built Snapchat's Discover personalization platform from the ground up and scaled personalized content systems from early product vision to 200M+ daily viewers
- Managed and scaled an 80+ person engineering org spanning backend infrastructure, ML/ranking, Android, iOS, and web engineering
- Partnered with product and leadership teams on recommendation systems, creator ecosystems, content quality, reliability, experimentation, and large-scale mobile infrastructure
What was technically hard
- Building a recommendation system that served hundreds of millions of users across wildly different content types (publisher stories, user stories, shows, editorial) with real-time personalization
- Scaling the engineering org from a small team to 80+ while maintaining quality and shipping velocity through Snapchat's most volatile growth period
- Balancing content quality and moderation at scale with the speed required by a social platform where content is ephemeral
What I learned
- How large-scale consumer recommendation systems actually work in production, not just in papers
- How to build and scale an engineering organization under hypergrowth pressure
- How content ecosystems, creator incentives, and platform dynamics shape technical decisions
- How to operate under constant market shifts while maintaining long-term technical bets
Why it matters now
The Snap chapter is where I learned to operate at scale. Recommendation systems, content platforms, and personalization at 200M+ daily viewers gave me intuition for how AI-native products need to work in the real world. That experience directly informs how I help founders think about product-quality loops, ranking systems, and scaling engineering teams.