At Bharat Electronics Limited, I developed a machine learning-based predictive maintenance system to analyze equipment sensor data and identify potential failure risks. I also built a Streamlit dashboard to show equipment health, predictions, feature importance, and sensor trends.
At SIRISOFT SOLUTIONS PVT. LTD., I developed an AI-powered invoice extraction application using LLM and OCR techniques. I integrated it into a Flask application and stored extracted invoice data in SQLite for search and retrieval.
As a final-year B.E. Computer Science and Design student at the University of Mysore, I’ve also built a PyTorch training framework with GPU profiling, mixed precision, and checkpointing. I’m developing adaptive optimization rules to improve training efficiency and reduce memory use.
