ML Engineer · Backend Developer · AI Builder
KIIT University, CSE | Bhubaneswar, India
- ML Researcher — building Generative AI pipelines with LangChain & FastAPI
- I build end-to-end AI products: from raw data → ML pipeline → deployed app
- Interested in LLMs, RAG systems, backend APIs, and real-world ML deployment
- Currently exploring distributed systems, Redis caching, and advanced NLP
- Certified: Data Analysis with Python (IBM) · Neural Networks & Deep Learning (DeepLearning.AI)
Languages
ML & AI
Backend & APIs
DevOps & Tools
| Project | What it does | Stack |
|---|---|---|
| PolicyLens | End-to-end RAG system that parses insurance policies and answers nuanced queries using Groq — FastAPI backend, React frontend, FAISS vector store | FastAPI · LangChain · Groq · FAISS · React · Supabase |
| LLM Semantic Cache | Semantic caching layer for LLM APIs — reduces redundant calls by embedding query similarity lookup against a Redis vector store before hitting the model | FastAPI · Redis · FAISS · OpenAI · Python |
| C++ P2P Network | Peer-to-peer file sharing system built from scratch in C++ — custom socket layer, node discovery, chunked transfer protocol | C++ · Sockets · TCP/IP · Multithreading |
| Transformer From Scratch | Full PyTorch implementation of "Attention is All You Need" — multi-head self-attention, positional encoding, encoder-decoder stack built from first principles | PyTorch · Self-Attention · NLP |
| Real Estate AI Suite | End-to-end ML platform — price prediction (R²=0.92), analytics dashboard, property recommender | Scikit-learn · Streamlit · Plotly · Docker |
"Build things that work in the real world, not just in notebooks."
