At Morgan Stanley, I built enterprise AI systems for trade operations, including an MCP-powered command layer that reduced manual support workflows from 30 minutes to minutes.
I developed a GPT-4.1 RAG assistant for trade-enrichment and exception dashboards, designing semantic retrieval, prompt orchestration, missing-field recovery, and tool-call guardrails.
I also automated legacy Java/XML migrations into Spring Boot and Angular workflows, led a testing migration to Vitest that raised coverage from 0% to 90%+, and built Playwright MCP automation for natural-language QA testing.
Beyond Morgan Stanley, I build applied AI products such as BoardPrepAI and an automated proctoring system, and have won national hackathons including Morgan Stanley Code to Give, D-Spark AI Challenge, and Campus Shark Tank.
