At London Stock Exchange Group (LSEG), I built an enterprise GenAI-powered document intelligence platform that delivered £350K+ in annual OPEX savings and reduced manual effort by 50 FTEs.
I also designed a context-aware RAG Answer Engine with source-grounded responses, conversational memory, and retrieval evaluation. I developed a LangGraph-based multi-agent proof of concept for company fundamentals analysis, which was demonstrated to regional executives and helped shape the team’s AI roadmap.
Earlier at LSEG, I built a regulatory extraction pipeline for Korean market filings that cut time-to-market from 15 days to 40 minutes across 130K+ documents. My work also included hybrid ML and rules-based ingestion, plus document intelligence pipelines for company fundamentals.
Before LSEG, I worked as a Reconciliation Analyst II at State Street, directing daily complex data reconciliations for 100+ EMEA-region entities and developing PowerBI exception-trend dashboards. I also built projects including an LLM-powered YouTube note-taking Chrome extension and an open-source terminal-first investment research tool, FinScope.
