Pranjal Shandilya

I've built AI-powered analytics and research applications using FastAPI, LangChain, Groq, React, Docker, MongoDB, PostgreSQL, Pinecone, and Supabase.

In an end-to-end e-commerce analytics platform, I processed 100K+ transactions across nine datasets, trained Random Forest models for churn prediction and revenue forecasting with a 77% R² score, and deployed containerized frontend and backend services through CI/CD. I also built a RAG research assistant that supports document uploads, source-cited answers, semantic search, and multi-document agent reasoning.

At IIRS, I developed geospatial analysis pipelines using Google Earth Engine and Sentinel-2 imagery for LULC classification, urban-sprawl detection, and change analysis. My DRDO internship also gave me exposure to defence-oriented software research, secure development practices, documentation, and engineering workflows.