Back to events

Turbocharge Tomorrow · Episode 11

AI Agents in Production: What Actually Breaks

Real incidents, real fixes: a candid look at running autonomous agents at scale, from eval drift to guardrail failures.

About this episode

Shipping an agent demo is easy. Keeping it reliable in front of real users, real data, and real edge cases is the hard part. In this episode, we sit down with engineers who have run agentic systems in production to talk about what actually breaks.

We'll cover eval drift after model upgrades, guardrail failures under adversarial input, and the observability gaps that let small issues become incidents.

Bring your questions — the last 15 minutes are reserved for live Q&A with the panel.

Speakers

  • Emanuel Duarte

    Head of AI Engineering, Distillery

Virtual Webinar

Link and calendar invite are sent by email right after you register.

Make moves with Distillery.

Choosing the right nearshore engineer or agile expert doesn't just save time — it transforms the trajectory of your entire team. Let's discuss your project today.

Get in touch