M.Tech student at IIT Kanpur working at the intersection of data science, operations research, and generative AI. I build systems that go beyond fitting data — I care about models that generalize in production.
- 🔭 Recently: Data Science Intern @ Centre for Energy Regulation — multi-model electricity demand forecasting across 33 Indian states
- 🧠 Interests: Time-Series Forecasting · NLP · GenAI/RAG · Combinatorial Optimization · Deep Learning
- 🎯 IIT Kanpur M.Tech (IME)
ML / DL
NLP / GenAI
Data & Optimization
| Project | Domain | Key Result |
|---|---|---|
| 🚛 Vehicle Routing with Installation Tasks | Operations Research · MILP · CPLEX | Exact solutions for 15 benchmark instances |
| 🤖 Offline Medical RAG Chatbot | GenAI · LangChain · FAISS · TinyLlama | Context-grounded Q&A with source citations |
| 🧠 Brain Tumor Detection | CNN · Transfer Learning · InceptionV3 | 90% accuracy on 4-class MRI dataset |
| 📊 Customer Sentiment Analysis | BERT · LSTM · NLP | 90% accuracy with fine-tuned BERT |
| 📈 NVIDIA Stock Prediction | LSTM · SARIMA · Time Series | LSTM RMSE = 11.4, outperformed all baselines |
Data Science Intern · Centre for Energy Regulation (May 2025 – July 2025)
Built multi-model electricity demand forecasting system (OLS, ARIMAX, Ridge, GM(1,1)) across 33 Indian states. ARIMAX identified as best production model over OLS on forward-looking scenarios.
"Models that fit history are common. Models that generalize the future are rare."