Talk

Virtual

Humans on the loop, not in it: Taking agentic engineering end to end

Using AI to generate code isn't enough. Platform engineers can enable teams to put agents to work across the full delivery loop: specify, build, test, release, run, and improve. Humans steer outcomes while agents do the heavy lifting.

CEST

Providing tools for developers to write code is not enough. In this talk, Kief Morris explains how agents can support the end-to-end software delivery process into production and keep improving it.

Key points:
• Teams should focus on defining outcomes and prioritizing improvements, not managing every artifact; humans should stay on the loop, not in it.
• Agents should do the heavy lifting: managing specifications, building and testing, assessing outputs at each stage, and implementing the improvements the team sets.
• Platform engineers need to enable the capabilities that make end-to-end agentic workflows reliable, especially for release and operations.
• As feedback loops tighten across the full cycle, teams progressively trust agents with more.

Virtual

Register for PlatformCon 2026