I've built agentic AI products including StateStack, an AWS Bedrock and Lambda-powered learning assistant with personalized learning paths, conversational state handling, tool calling, and context retrieval.
At Utah Tech Labs, I validate production-scale multimodal datasets, create QA workflows, and work with ML engineers to resolve edge cases, anomalies, and model-ingestion blockers.
During my IIT Kharagpur research internship, I developed biomedical image-processing workflows using Attention U-Net and Monte Carlo simulations, while migrating legacy MATLAB work to modular Python pipelines on cloud GPU runners.
I've also shipped an AI-driven sales and profit analyzer at Infosys Springboard and built REST APIs, dashboards, predictive models, and cloud-deployed applications using Python, React, Flask, Docker, AWS, GCP, MySQL, and MongoDB.

