- Analyst and consultant to MSEDCL, India's largest Electricity Discom.- Developed and Deployed real-time intraday forecasting for state's electricity demand using Machine Learning algorithms like Random Forest, XGBoost, combined with RNNs, LSTMs and time-series models, with overall error of 1.2%.- Worked on Day ahead, weekly, long-term and renewables(wind and solar) forecasting.- API development for real time SCADA demand fetching.- Models are currently being used by the discom for resource planning and minimizing losses.

Taught Web Development, Python, Machine Learning and Mathematics to more than 50 students, worked on collaboration with local schools, startups and institutes.

- Worked on Day-ahead and intraday forecasts for 7 Discom of 4 states using Time Series Analysis and Machine Learning techniques, improving the overall MAPE from 7-8% to 2-3%.- Paper (Unpublished): Machine Learning Based approach for Demand Forecasting in Himachal Region

- Worked on the main product Vayam (Indigenous Video Conferencing Platform) on improving User Experience, Growth strategies, Product improvements and Analytics- worked on CRMs and marketing tools like MixPanel Analytics

Organized Think India Startup and Innovation E-Summit 2020.