Deployed a scalable AI solution using Python, Azure Machine Learning and Azure AI Services to automate unstructured handwritten text extraction for claim processing. Reduced operational costs by 90% by replacing a team of 40 manual validators and decreased processing time to under one minute per document. Achieved 98% precision across more than 10,000 documents, demonstrating system scalability and reliability in production environments.

Involved in developing a POC for the prediction of train shoe brake thickness. Developed a POC for real-time ship detection in surveillance feeds using YOLO.