Developing Firmwide GenAI

Going through the Technology Analyst Program (TAP); an 8-week training program, on the Application Development track.Worked in a team of 4 to modernize an existing application with multiple services at Morgan Stanley in 3 days. Upgraded Angular, replaced SOAP with RESTful services using Spring Boot, improved existing messaging services using Kafka and Zookeeper, and containerized each service individually using Docker.

• Developed a Proof-of-Concept (POC) for conversationalizing and orchestrating tools used by ~16,000 Financial Advisors into the AI @ Morgan Stanley Assistant chatbot. Utilized Flask, GPT 3.5/4, and OpenAI Function Calling in Python.• Achieved a 30% reduction in POC development time by engineering a chat interface featuring document chat, user session management, multi-modal responses, and a uniform REST API using Angular in TS.• Contributed to the Architecture Review Process for the conversationalization of tools using internal APIs.• Played a key role in Generative AI vendor evaluation strategy by documenting vendor capabilities and product features, developing standardized classifications and KPIs. Participated in product demos and conducted POC testing.

• Increased consumer servicing capacity by >100% automating the consumer onboarding process through an automated form and providing consumer API service usage monitoring through a dashboard interface with parameterised search.• Created the database schema by documenting requirements through meetings with Business Analysts and Managers.• Utilised React-TypeScript for Frontend, Java-Spring boot for Backend, and MS SQL Server for database. Deployed using TeamCity for CI/CD and integrated SonarQube for code coverage and quality test.