At Tata Consultancy Services, I build AI systems that turn requirements into code, generate ADAS test cases from NCAP documents, and automate code review. My AWS Bedrock, AgentCore, and serverless RAG work reduced documentation effort by 90% and accelerated feature turnaround by 75%.
I’ve also developed computer vision systems for railway inspection using DeepLabV3, PatchCore, and Stable Diffusion across multi-camera pipelines processing 20K+ images per cycle. My Azure MLOps pipeline for YOLO-based traffic sign detection delivered roughly 30 FPS inference and over 90% validation accuracy.
Across Django, React, Node.js, and machine learning projects, I’ve built recommendation, accessibility, and full-stack applications with a focus on practical impact.
