Evolving the Tree: Multi-Agent LLMs that Plan, Retrieve, and Ship
Details
The Business Problem
In the real world, AI solutions work crosses documents, systems, and approvals. You need agents that can plan, look things up, challenge each other, and show their work with clear guardrails.
The Tools
Lightweight orchestration that explores options before acting
Retrieval and memory so answers are grounded in your corpus
The Takeaways
A copy-able reference architecture for multi-agent work
A starter path - add retrieval, add a critic, add review, then instrument quality
Agenda:
05:30 - 05:55 Welcome & Networking
05:55 - 06:00 Announcements
06:00 - 07:00 Main presentation
07:00 - 07:15 Q&A
07:15 - 07:30 Wrap-up
Speaker: Gaja Kannan
Builder | Learner | Business Technologist
Head of Architecture and AI Enablement @ MSIG USA
Applying and showcasing use cases in insurance segment
Sponsor: Ippon Technologies
Parking available on both sides of the building and on the street.