

What Just Happened?! Observability for Large-Scale Systems and Agents
AI makes it much easier to build software. Does it also make software easier to understand, test and run reliably?
Speakers:
Liz Fong-Jones (Technical Fellow, Honeycomb)
Laurie Voss
Qian Li (co-founder and CEO at DBOS) as a moderator
What we’ll cover:
Observability for AI: How do we understand what agents are actually doing and what would we design differently in today’s observability stack?
AI for observability: Can agents reliably investigate incidents, identify root causes, and test hypotheses against real production data and what does it take to make observability data understandable and useful to agents?
Testing in the AI era: When code is cheap to write, who verifies it? Are traditional tests enough, or do we need evals, traces, and production feedback?