Why product managers should want their AI agents to fail loudly

Uxcel on YouTubeWatch on YouTube

SaveMark complete

Sign in to save videos and track what you’ve finished.

IntermediateUX51 min

About this video

A conversation with Daniel Zherenkov, Senior Product Manager for AI and ML, on running a team of AI agents around a real product and deciding which calls stay with a human. It covers guardrails, agent memory and choosing models for planning and building.

What you’ll learn

  • Set up guardrails that keep a human in charge of agent decisions
  • Structure a small team of AI agents around one product
  • Split agent memory into layers and keep the files usable