As organizations race to build AI agents, many teams are encountering challenges that feel new: evaluation uncertainty, feedback loops, behavioral drift, exploration versus exploitation, and maintaining user trust in systems that continuously adapt.

But these challenges are not new.

Large-scale recommendation and personalization systems have spent years operating as adaptive systems in production, continuously learning from user behavior while balancing relevance, business objectives, latency constraints, and long-term system health.

In this talk, we'll examine the architectural and operational lessons learned from building adaptive recommendation systems and explore how those lessons apply directly to modern agentic architectures. We'll cover feedback loops, evaluation frameworks, observability, adaptation strategies, and organizational structures that enable continuous learning without sacrificing reliability.

Rather than focusing on model architectures, this session focuses on the systems that surround intelligence—and why they ultimately determine success in production.

Key Takeaways:

1. Understand why recommendation systems and AI agents share many of the same architectural challenges.
2. Learn practical approaches for operating adaptive systems safely in production.
3. Explore evaluation strategies for systems whose behavior changes over time.
4. Understand feedback loops, behavioral drift, and exploration-exploitation tradeoffs.
5. Learn organizational patterns that help teams build trustworthy adaptive systems at scale.