Talks

Lessons learned from building autonomous agent teams with an emphasis on why reliability is the lens for evaluating quality over time.

Kam walks through production scenarios, the unpredictability inherent in agent systems, and the frameworks he uses to align teams on measurable quality targets.

A step-by-step build of the smallest useful agent, showing how planning, memory, tool use, and control flow combine to create an end-to-end system.

The talk demystifies modern hype by incrementally layering capabilities so engineers can see exactly how each choice affects the whole agent.

A candid look at the culture shift required to share and learn from production outages instead of hiding them behind “if you just” explanations.

The session encourages engineering teams to normalize post-incident storytelling so everyone can build safer systems together.

An extended failure report that shares real serverless missteps so the broader community can level up without repeating the same mistakes.

By unpacking the root causes, Kam shows how embracing honest retrospectives leads to better engineering instincts and more resilient platforms.