At Centric Consulting, I built asynchronous Python services that expose LLM capabilities through authenticated APIs, and designed RAG pipelines over code and documents using Claude and OpenAI SDKs.
I also built a code review agent that uses retrieved code context and Claude Sonnet to analyse pull-request diffs, flag issues and propose fixes. I established prompt evaluation sets and regression checks, and led a five-member AI team delivering proofs of concept.
Previously, at Optum Global Services, I led a 12-engineer team migrating a healthcare analytics platform to Azure microservices. My work has also included telecom platforms at Ericsson and Comptel, personal projects such as Knowledge Crib and Seero Flow, and two research manuscripts with Springer Nature, publication pending.

