Watch all of our recorded sessions from Interrupt here: https://interrupt.langchain.com/video/?utm_medium=social&utm_source=youtube&utm_campaign=q2-2025_interrupt-2025_co
David Odomirok and Zheng Xue from JP Morgan Chase Private Bank reveal how they built "Ask David" - a sophisticated multi-agent AI system designed to automate investment research for thousands of financial products. With billions of dollars in assets at stake, this isn't just another chatbot - it's an enterprise system built with human oversight for high-stakes financial decisions.
What You'll Discover:
- The exact multi-agent architecture JP Morgan uses: supervisor agents, specialized sub-agents, and human-in-the-loop workflows
- How they evolved from simple ReAct agents to complex sub-graph systems through iterative development
- Their evaluation-driven approach that improved accuracy from 50% to 90%+ for domain-specific queries
- Why the "last mile" from 90% to 100% accuracy requires human SME oversight in high-stakes environments
- Real-world lessons on structured vs. unstructured data handling in enterprise AI
#LangChain #MultiAgent #LangGraph