I build Android apps, write about machine learning, and create resources for developers who want to ship ML-powered products.
Mobile apps designed to solve real problems with elegant, intuitive solutions.
Professional teleprompter app for content creators, speakers, and video producers. Features smooth scrolling, customizable display, floating window mode, and more.
Practical guides for developers who learn by building real projects.
Build FinRisk — a complete credit risk assessment app running a trained ML model entirely on-device. No server, no API calls, no internet required. Train in Python, convert to TFLite, run inference in under 15ms on Android with Clean Architecture + Hilt + Compose.
Thoughts on Android development, machine learning, and building real products.
The story behind building a credit risk classifier that runs on-device, and why I wrote a book to teach other Android engineers to do the same.
Read article → Towards AITaking deep learning theory from Stanford's CS230 and applying it to a production Android app with on-device inference.
Read article → ProAndroidDevA deep dive into Android Studio's Journeys feature and how to scale it from simple demos to enterprise-grade test coverage.
Read article → MediumA practical 4-part series on modularizing Android apps, from identifying module boundaries to migrating shared code.
Read series →I'm a software engineer specializing in Android development and on-device machine learning. I build apps, write technical books, and publish articles to help developers ship ML-powered products.
My work focuses on making machine learning practical and accessible for mobile developers — no PhD required. I believe the best way to learn is by building real projects, not reading papers.
When I'm not coding, I'm writing about Android architecture, modularization, and the intersection of ML and mobile development.
Have questions, feedback, or want to collaborate? I'd love to hear from you.
sayvortana.itc@gmail.com