About me
I'm Ashish Mehrotra, an Android developer who got pulled deep into the AI rabbit hole and never looked back. I spend most of my time building mobile apps and experimenting with running AI on-device, building agents, and figuring out what all this new tech actually means in practice.
Over the last year I've built a bunch of PoCs: LLMs running locally on Android, agents using ADK and AWS, Jetpack Compose apps integrated with Gemini, and more. This site is where I document all of that: what I built, how I built it, and why.
I believe the best way to understand new technology is to build something with it. So whenever something new drops (a new model, a new framework, a new technique), I try it, break it, and write about it here.
What I work with
Mobile
- Android
- Kotlin
- Jetpack Compose
- Coroutines
- MVVM
AI / ML
- On-device LLMs
- GGUF / llama.cpp
- AI Agents
- ADK
- Gemini API
Tools
- Android Studio
- Git
- Python
- Gradle
- Firebase
Right now
Building AI-powered Android apps and agents
Exploring on-device ML with GGUF, llama.cpp, and MediaPipe
Writing about what I learn, because the AI space moves too fast not to
Sharing what I build and learn on this site and on LinkedIn