Android apps and AI tools — with a focus on privacy, clean UX, and no subscriptions.
Started coding in 2025. Built Anti-Gravity from scratch — React Native, Firebase, zero-trust auth, Play Integrity API, and server-side billing. Learned by shipping, not by following tutorials.
I care about one thing: software that works for users, not against them. No ads, no subscriptions, no data collection.
Privacy-centric PDF editor solving top market complaints — no data collection, no ads, no subscriptions. Engineered zero-trust local auth & Google Play Billing verification.
Try App
Autonomous narrative detection for the Solana ecosystem. 14-day signal synthesis with high-novelty extraction.
Autonomous job application agent using Gemini & n8n. Intelligent discovery, scoring, and resume tailoring.
Turn any text into study tools. AI-powered mindmaps, flashcards, and quizzes to accelerate your learning.
Farcaster/Base mini-app streamlining onboarding and ecosystem navigation.
Internal analytics suite for Anti-Gravity. Tracks DAU/WAU/MAU, tool error rates, conversion funnel, retention, and real-time user behaviour — built from scratch on raw Supabase events.
Things I've actually shipped with, not just read about.
Powers the AI redaction engine in Anti-Gravity and the job scoring in CareerPilot. Used for real document understanding, not demos.
What Anti-Gravity is built with. One codebase, live on the Play Store.
Used for MindMint and CareerPilot AI. Handles both the UI and the backend API routes.
Play Billing API, Play Integrity API, and server-side purchase verification. Built the full billing stack for Anti-Gravity — five layers of fault tolerance so no purchase gets lost.
Auth, Firestore, and Functions across most projects. Supabase for Anti-Gravity's backend.
Built CareerPilot AI's automation pipeline with it — job discovery, scoring, and resume tailoring without writing all the glue code manually.
Tested on Berachain, Base, Walrus (Sui), and Story Protocol. Know how these ecosystems actually work at a user and builder level.
No team, no mentor. These are real production bugs I debugged and fixed on my own.
After launching paid access in Anti-Gravity, every purchase was being rejected server-side — but users had no idea why. The rate limiter was configured as fail-secure, which blocked all billing verification requests. Real users were paying and getting nothing. Switched to fail-open design. Success rate went from 0% to 99.9% overnight.
After adding Google sign-in, some users who had already purchased premium were being downgraded back to free tier. The bug came from two sources of truth — local storage and the server — disagreeing after login. Fixed it by building a handshake reconciliation step that runs on every auth event and settles the correct tier before the app loads.
Android Not Responding reports started appearing in Play Console. Spent hours tracing it — turned out the freeze was coming from Google's own license verification library, not my code. I couldn't fix what I didn't write, so I restructured how and when that code runs to avoid the blocking call on the main thread.
Building while learning. Always.
Learning Python daily — focused on AI engineering applications. Data pipelines, backend scripting, and integrating with LLM APIs. Building alongside learning, not waiting until it's perfect.
Moving from "AI user" to "AI engineer." Studying agentic systems, prompt architecture, and how production AI pipelines actually work under the hood — not just calling APIs.
Looking to integrate AI or build a mobile product? I'm available for new partnerships.
