Portfolio
Full Stack & AI/ML Developer — building systems where intelligence meets real-world impact
I'm a B.Tech IT student at Dr. N.G.P. Institute of Technology, Coimbatore (CGPA 8.02), with a dual identity as both a Full Stack and AI/ML developer. I've builtsystems ranging from distributed job queues on AWS to deep learning diagnostics for medical imaging.
My passion lives at the intersection of clean architecture and intelligent systems whether that's a RAG pipeline answering questions over PDFs, or an ensemble CNN diagnosing brain tumors with Grad-CAM explainability.
Multi-class MRI classification using Transfer Learning on ResNet18, DenseNet121 & MobileNetV2 — trained on 5,600 images across 4 tumor classes with soft-voting ensemble.
Distributed job queue with decoupled producer-consumer architecture, visibility timeouts, dead-letter queues, and at-least-once processing guarantees.
Fully local PDF-based RAG system — extraction → chunking → MiniLM embeddings → FAISS similarity search → context-aware QA with FLAN-T5.
Peer-reviewed paper presented at ICECA-2025 — a premier forum for research in electronics, communication and aerospace engineering. Explores RFID-based automation for frictionless retail checkout, contributing to the conference's focus on AI and smart systems.
Open to full-time opportunities, collaborations, and interesting conversations about AI, systems, and everything in between.
Python
React
Next.js
FastAPI
PyTorch
Docker
AWS
PostgreSQL
MongoDB
TypeScript
JavaScript
Git