AI Engineer · Full Stack Python Developer · Data Pipeline Builder
I build production-grade AI systems — from LangGraph stateful agents and RAG pipelines to real-time Kafka/Spark data infrastructure. I deploy things that actually work.
I'm a Full Stack Python Developer and AI Engineer focused on building systems that sit at the intersection of backend engineering, machine learning, and agentic AI.
My flagship project — a LangGraph-powered GitHub PR Review Agent — uses Groq + LLaMA 3.1 70B, pgvector RAG, and GitHub webhooks, deployed live on GCP Cloud Run. It reviews code like a senior engineer, automatically.
Beyond AI agents, I build real-time data pipelines with Kafka, Spark, and Airflow, and full-stack apps with FastAPI backends and React frontends. I care about systems that are production-ready, not just demo-ready.
An AI agent that reviews GitHub Pull Requests like a senior engineer. Triggered via webhooks, it fetches diffs, retrieves codebase context via RAG, and posts structured review comments automatically.
End-to-end streaming pipeline predicting NYC taxi demand in real time with Kafka, Spark, and ML — including scheduled Airflow retraining.
Full-stack loan management platform with JWT auth, ML-based risk scoring, admin dashboard, and real-time installment calculation.
CNN-powered Flask web app classifying MRI scans as Tumor / No Tumor with OpenCV preprocessing and MongoDB Atlas storage.
Fine-tuned ResNet-34 on plant disease dataset using PyTorch. Flask app for real-time leaf image classification.
Chat with any PDF using LangChain and vector search. Extracts, chunks, embeds, and answers context-aware queries in real time.
A Three.js particle physics simulation. Add particles, watch them orbit and collide — generating plasma explosions with audio feedback. Built for fun.
Let's build
something real.
Open to AI/ML, backend, or data engineering roles. Also happy to collaborate on interesting projects.
{
"available": true,
"open_to": [
"AI/ML Engineer",
"Backend Developer",
"Data Engineer"
],
"location": "Delhi, India",
"remote": true,
"response_time": "< 24 hours"
}