At Northern Trust, I architected an enterprise RAG system using OpenAI APIs, LangChain, and Pgvector/Redis. It cut research retrieval time across compliance documents from 25 minutes to under 15 seconds.
I also integrated LLM output validation and prompt guardrails into CI/CD pipelines, reducing hallucinations in automated client reporting by 35% while meeting SOC 2 data security requirements. My work there includes backend microservices and event-driven pipelines for financial data and transaction logs.
Previously, at DXC Technologies, I built cloud-native microservices and API gateways on Microsoft Azure, and set up automated deployment pipelines. In personal projects, I developed a RAG knowledge engine for enterprise documents and an asynchronous task orchestrator for LLM inference.

