Building Androidify: an AI-powered Android experience
summary
An open source Android app converts a user photo into an Android bot using Firebase AI Logic SDK plus ML Kit and a custom fine‑tuned imagine model; image validation runs on Gemini 2.5 Flash, description and image gen use Firebase but fall back to Gemini Nano on-device when available, cutting three server calls to one and keeping photos local. The imagine model is fine‑tuned from over a thousand bot variations, uses ML Kit segmentation for stickers, and adapts across foldables, tablets, XR and watch faces.
AI-written summary · about
chapters (16)
- 0:00Welcome to Androidify
- 2:51Live demo: Rob to Android bot
- 6:03AI-powered features overview
- 6:47AI image processing pipeline
- 9:35Adding vibes and backgrounds
- 10:59Firebase AI Logic best practices
- 12:01On-device AI with Gemini Nano
- 14:47Sticker mode with ML Kit
- 15:52Delightful UX design
- 17:14Expressive animations with Material 3
- 21:26Adaptive layouts for all screens
- 25:24Custom previews for development
- 27:11Foldables and rear camera APIs
- 31:30Expanding to XR and watch faces
- 32:31Summary and try Androidify
- 33:36Q&A on image generation
related talks
30mGenAI on Android with Gemini Nano and Firebase
39mAdd Generative AI capabilities to your Android app
41mBuild intelligent Android apps with Google's AI
41mGen AI for Android Developers
21mAdd Generative AI capabilities to your Android app
01h 11mBuild Your Own AI Sidekick with Flutter
39mGetting Started with GenAI for Android Apps
10mAI in your pocket: Building intelligent Android apps