At Amazon, I build evaluation pipelines for agentic AI workflows, testing multi-step reasoning, tool-calling accuracy, and prompt adherence across production samples. I also analyze failures in Supervised Fine-Tuning workflows and define corrective alternatives.
I defined quantitative quality gates and regression benchmarks at Amazon, achieving a 96% quality score while addressing data-quality and reliability gaps. Before that, as an Associate Technical Consultant at Adobe, I developed and shipped 20+ responsive frontend components and optimized performance to reach Lighthouse scores of 95+.
At Tata Consultancy Services, I used Python and SQL to investigate transactional data and support production applications. As a Software Development Intern at AB InBev, I built internal web tools that cut manual process time by 60%; I also worked on SplitUp, an expense-splitting app with analytics dashboards.
