CS (Cyber Security) undergrad Β· AI/ML Engineer Β· Samsung R&D Intern
I build multimodal vision-language models, full-stack AI platforms, and security tooling β end to end.
- π B.E. Computer Science & Engineering (Cyber Security) @ MS Ramaiah Institute of Technology β CGPA 9.53 / 10
- π¬ Project Intern @ Samsung PRISM / Samsung R&D Institute India β building EmoCapNet, a 254.8M-param multimodal VLM
- π Co-author of a Springer Nature book chapter on Quantum Computing for IoT
- π LeetCode 1821 β top 7.32% globally Β· π₯ Campus Ambassador, Kshitij (IIT Kharagpur)
- β‘ I care about correctness, measurement, and things that actually ship.
| Project | Description | Stack |
|---|---|---|
| π Smart Course Generator | Full-stack AI platform that generates multi-module courses streamed over SSE, with adaptive quizzes, flashcards & AI-scored mock interviews. Custom multi-provider AI router with circuit breakers + LLM-as-judge evals in CI. | React Β· Node Β· MongoDB Β· SSE |
| π§© Rubik's Cube Studio | Kociemba's two-phase solver written from scratch β averages ~20.6 moves (God's number is 20) at 100% correctness over 1,500+ cubes. Web Worker + Three.js 3D animation with webcam scanning. | TypeScript Β· Three.js Β· Vitest |
| π‘οΈ DNSentinel | DNS threat-intelligence platform detecting DGA, tunneling, exfiltration & C2 beaconing via an RF + Isolation Forest ensemble, with a SOAR auto-block layer mapped to MITRE ATT&CK. | Python Β· FastAPI Β· scikit-learn |
"Correctness, measurement, and things that actually run."

