Work
Seven years of Android, most of it on messaging used by hundreds of millions of people.
Where I’ve worked
- 2025 – now
Senior Software Engineer, Android
· BengaluruLeading a team of three building on-device fraud detection, which flags 20M+ fraud messages a day across India and is rolling out to new markets. Proposed and built the on-device ML platform behind it (BERT-style models at 20ms p95), and a shared Kotlin pre-processing library that lifted agreement between the phone and data science’s pipeline from ~80% to ~99%. Proposed the remote rules engine behind 500M+ Message IDs a day, which product, ads and enterprise teams change without a release, and led the app’s move to Material 3.
- 2021 – 2025
Software Engineer, Android
· BengaluruCo-built Message ID, a Caller ID-style SMS overlay drawn with Compose straight through WindowManager, at 350M+ daily impressions; then built the infrastructure for its LLM summaries, made without any message content leaving the phone, which took it to 500M+ a day. Scaled Smart SMS from 1–2 to 15+ business partners, sped up app startup on every incoming SMS by cutting parser init from 700ms to 450ms, and shipped Truecaller’s first production Jetpack Compose screen.
- 2019 – 2021
Software Engineer, Android TV
· NagpurOwned the player UI and parental controls for Verizon’s Stream TV launcher. Made trick-play thumbnails render 60% faster by prefetching them in the background based on how people seek, and cut startup CPU overhead by 30%.
- 2018
Web Development Intern
WoodFlake· IndoreEvent-driven Node.js APIs and a Vue.js admin dashboard, deployed on AWS.
Hackathons
Winner, Labdays 2023 & 2024
TruecallerInternal hackathon wins for WhatsApp Caller ID (2023) and Message ID animations (2024).
Winner, 30 Hacks 2019
GlobalLogicAn app that flags early cancer risk from lifestyle data.
Tools I reach for
- Languages
- Kotlin · Java · TypeScript · JavaScript · Bash
- Android
- Jetpack Compose · Coroutines · Hilt · Room · Canvas & AGSL shaders · Material 3 · Kotlin Multiplatform · Android TV
- On-device ML
- TFLite / LiteRT · Firebase ML · ML Kit · BERT pre-processing · Kotlin–Python interop (JPype, Py4J)
- Tools & testing
- Git · Gradle · Firebase · Android Studio Profiler · JProfiler · JUnit · MockK · Espresso