ACM Dallas, Dallas, TX25 min · 28 slides
Harness Engineering: How Reliable AI Agents Actually Get Built
The model is the engine; the harness is the car. Why most agent pilots stall on reliability rather than capability, and the engineering that closes the gap: tools and permissions, verification loops, guardrails, evals, and human handoff.
AI agentsReliabilityEvalsDeveloper tooling
Open deck