At Cogensus AI, Inc., I define expected AI behavior, build evaluation datasets and grading rubrics, and analyze recurring failures to prioritize product improvements. I also tune RAG and agentic AI workflows, testing changes against quality and performance criteria.
As an AI & LLM Engineer, I designed and deployed AI services for healthcare workflows, reducing clinical decision time by 45%. Earlier, at Phoenix Global, I developed machine learning services for insurance workflows that improved prediction accuracy by 25%; my research at GITAM University contributed to IEEE-published hearing-loss classification results.
