Logan Kilpatrick: Build and Ship AI Apps Fast

​You've got model access, the tools, and an idea. But somewhere between the first prototype and a real product, most builders lose momentum — they wait for the perfect plan instead of shipping something rough, or they polish one feature for a week instead of testing ten. The gap between "I could build that" and "I shipped that" is where most projects quietly die.

​In this live session, Logan Kilpatrick breaks down how he actually moves fast: how he takes initiative without waiting for permission or a perfect spec, how he uses iteration speed as an unfair advantage, and how he decides what's worth building versus what's just busywork. He'll get into the mindset that separates the builders who ship from the ones who stall — and how that same agency shows up in building AI products people actually want, not just AI demos that impress for a week.

​For context, Logan works on Google AI Studio and the Gemini API at Google DeepMind, leading the team focused on giving developers worldwide the tools to build with Gemini, Veo, Nano Banana, and more. Before Google, he led developer relations at OpenAI, worked as a machine learning engineer at Apple, and advised NASA on open-source policy. Few people have spent as much time in the room with builders shipping at the edge of what's possible with AI right now — this is a chance to hear directly how he thinks about speed, initiative, and building things people want.

​This session is part of the Build with Gemini XPRIZE — a 90-day hackathon in partnership with Google and XPRIZE challenging builders to create a real AI business from scratch. $2M in prizes across 25 winners, including a $500K grand prize. Submissions close August 17, 2026.

​Register to compete: https://www.geminixprize.com

​If you're a builder in the competition racing toward your submission and want to know how to move faster, take more initiative, and build something people actually want — this is the session that shows you how.