Most AI assistants today are static—they reset every time you talk to them. In this session, you’ll learn how to transform LLMs into adaptive, agentic systems that learn from context, memory, and experience.

We’ll walk through how to build a tool-enabled agent with memory, demonstrate how MCP (Model Context Protocol) and A2A (Agent-to-Agent) communication enable agents to collaborate, and show how they can integrate with real-world services like GitHub, Jira, or ServiceNow through MCP servers.
You’ll also see when it makes sense to fine-tune an LLM versus using memory-based continual learning so your agents can improve without retraining the model.

By the end of this session, you’ll understand the core patterns and trade-offs behind building adaptive, multi-agent AI systems that evolve over time—ready to scale from simple prototypes to production-grade architectures