name: Rupesh Bharambe location: Chh. Sambhajinagar, Maharashtra, India education: B.Tech AI & Data Science @ CSMSS CSCOE (2022-2026) incoming_role: Software Development Engineer @ RBI Innovation Hub (RBIH) current_role: Software Engineer Intern @ Biopan Scientific Pvt. Ltd. leadership: Vice President — AI & Data Science Students Association focus: AI/ML ×ばつ Public Infrastructure • Fintech • LLM/RAG • Full-Stack
Final-year AI & Data Science engineer joining RBI Innovation Hub as a Software Development Engineer. Winning NFPC 2026 — RBIH's own national challenge with 2000+ participants and ₹5L prize — was how this story started. I work at the intersection of applied ML and public-infrastructure problems: fraud detection, energy grids, satellite data, healthcare access.
| Competition | Award | What I Built | |
|---|---|---|---|
| 🏆 | NFPC 2026 | National Winner — RBIH ×ばつ IIT Delhi | Mule account detection — 0.999 AUC-ROC on 400M txns, ₹5L prize → SDE offer at RBIH |
| 🥇 | Smart India Hackathon 2024 | Grand Finale Winner | EDFS — AI electricity demand forecasting for Delhi's power grid |
| 🥇 | @scale AI Hackathon 2026 | Winner — Marathwada | AURA AI — 4 ML pipelines on real HVAC industry data |
| 🥇 | RACKATHON 2025 | Winner — GH Raisoni | Face Liveness Detection with custom neural architecture |
| 🥉 | ORCHATHON 2K25 | 3rd Prize — NK Orchid | InnoMediX — AI telemedicine platform |
| 🏅 | HackCelestial 2.0 | Top 5 — Pillai College, Mumbai | Satellite data platform |
| 🏅 | SIH 2025 Grand Finale | Top 5 | VETRA — Animal classification for Ministry of Fisheries |
| 🥈 | Avishkar 2024 | Silver — Zonal Level | Innovative AI research project |
| 🌍 | Pune Agri International | International Finalist | AgriTech AI platform |
6x Hackathon Winner · 15+ Finalist Finishes · Featured in 17 newspaper articles across 10+ publications
Incoming Software Development Engineer — RBI Innovation Hub (RBIH) Joining soon
Public infrastructure for India's financial system · ML for fraud detection, regulatory tech, and financial inclusion
Software Engineer Intern — Biopan Scientific Pvt. Ltd. Oct 2025 — Present
Desktop software for scientific instruments (LC, bioreactor, RT-qPCR) · 21 CFR Part 11 compliance · AI/ML analytics
Software Developer Intern — Intersense Technologies LLP Apr 2024 — May 2024
Raspberry Pi PyQt HMI · CNC compensation algorithms · SQLite traceability systems
Stacking ensemble (LightGBM + CatBoost + XGBoost) with 615 features, DuckDB streaming on 16GB RAM. Beat 2000+ participants. ₹5L prize → led to an SDE offer at RBIH.
7 ML models for Delhi power grid at 5-min/hourly/daily resolution. 93 engineered features, walk-forward CV. Deployed on Vercel + Render + Supabase.
477 facial landmarks + custom MLP for anti-spoofing. <200ms latency on low-resource devices. MediaPipe blendshape thresholds + offline classifier.
Knowledge Graph (Neo4j) + RAG + FAISS + Groq LLM for querying ISRO's MOSDAC satellite data. 12 satellites, 154 data products.
3 YOLO pose models + U-Net + ResNet18 udder classifier + ArUco scaling for NDDB-style trait scoring. Built for Ministry of Fisheries, Animal Husbandry & Dairying.
4 ML pipelines: demand forecasting (8.4% MAPE), anomaly detection, SciPy/Monte Carlo optimization (13ドルM profit uplift), market share NLP. Full-stack FastAPI + React dashboard.
AI & Machine Learning
Python TensorFlow PyTorch scikit-learn LightGBM XGBoost CatBoost OpenCV LangChain HuggingFace MLflow
Data Science
Pandas NumPy DuckDB Jupyter Apache Spark Power BI SQL
Backend & Databases
FastAPI Flask PostgreSQL MongoDB Redis Neo4j
Frontend
React Next.js TypeScript TailwindCSS Streamlit
Cloud & DevOps
Docker Kubernetes AWS GCP Vercel GitHub Actions Linux
Tools & Hardware