2nd-year CSE student at VIT Vellore. I build full-stack tools and ML/data systems around one rule: the project has to solve a real problem, not just demonstrate a tutorial.

• 🚀 I'm currently building AI-assisted systems for a national buildathon
• 🤝 I'm looking to collaborate on applied ML and intelligent data pipeline projects
• 🌱 I'm currently learning to scale ML models beyond notebooks into production-ready services

Ripple — A developer tool that shows how far a code change can travel through a software system. Combines static dependency analysis, runtime intelligence, deterministic risk scoring, architecture drift detection, and AI explanations, with CLI and GitHub Actions integration for pull-request risk checks. Built for the LatentForce BuildSprint 2026. Python FastAPI React TypeScript NetworkX OpenTelemetry GitHub Actions

driftradar — An AI-assisted monitoring and recovery system for developer data pipelines. Detects extraction drift, diagnoses failures, proposes a recovery via Bright Data Scraper Studio, and verifies the fix actually worked before closing the incident. Built for the WeMakeDevs × Bright Data Scrape-Verse Hackathon. React TypeScript Node.js Bright Data

tally-erp-automation — Local-first accounting tool that converts bank statement PDFs into Tally ERP 9-importable XML vouchers, with a review workflow, ledger mapping, and Excel export. Built for accountants and finance teams who currently do this by hand. Next.js FastAPI SQLite pdfplumber

Result_Management_System — A student results system with separate teacher/student portals, audit logging on every record change, bcrypt-secured auth, and auto-generated PDF marksheets. Python SQLite ReportLab

AI coding assistants are sometimes used in my projects for debugging, refactoring, and exploring approaches — always reviewed and understood before being shipped. Disclosed per-project where relevant.