Motivation Kaksha
Live50k+ active users
JEE college prediction platform serving 50,000+ active users. Built a high-performance REST API with a Redis caching layer to handle concurrent prediction requests at scale. Mobile app in progress.
A Software Engineer.
I build performant, production-grade systems usingTypeScript,React,Next.js,Rust andPostgreSQL. Focused on distributed systems and low-level architecture.
50k+ active users
JEE college prediction platform serving 50,000+ active users. Built a high-performance REST API with a Redis caching layer to handle concurrent prediction requests at scale. Mobile app in progress.
Behavioral email automation
Loops.so — but with intent scoring. Tracks user behavior, builds a behavioral fingerprint per user, and sends emails from your real inbox only when the algo decides the timing is right. Anti-spam scoring prevents over-mailing; goes beyond pure event triggers.
Anonymous recovery system
Lost-item recovery via QR codes. Finder scans, sends a message, owner gets notified — no personal info exposed on either side. Privacy-first architecture with zero account requirement for finders.
Self-hosted music streaming
Full music streaming pipeline: JioSaavn URL resolver + yt-dlp fallback feed audio into Cloudflare R2. A Worker API handles catalog, search, and range-aware streaming. Manifest-driven — no database needed.
Educational assistant
RAG-based educational assistant. Explains complex problems step-by-step using context-aware LLM agents with retrieval over course material.