My talk is called “Ordering Coffee with Firebase AI”, and it’s about a real case of bootstrapping AI assistants with Firebase AI Logic. I’m working on a coffee shop chain app called Drinkit. Our menu is huge: you can customize drinks with or without milk, change ice, add syrups, literally millions of combinations. Too big for a static UI. So we thought: why not let AI guide the user? What if you could just say: “It’s hot today, give me something refreshing”. And boom, the app recommends the perfect drink for you. That’s exactly what we’re building in Drinkit. I believe this talk will be useful because it’s a real production case of applying Firebase AI, not just sandbox examples. Attendees will see how we combined speech-to-text, Firebase AI models, and function calls to handle a huge menu (2M+ token context!) and deliver an interactive, personalized ordering experience. Developers will walk away with inspiration, code examples, and practical lessons on how to quickly bootstrap AI features in their own apps.