AI agents increasingly invoke APIs, access sensitive data, and take actions with real consequences. Governance cannot remain a document reviewed only after deployment. This session presents a practical two-layer approach: build-time tests for policy compliance, answer grounding, and tool-use trajectories, followed by runtime enforcement using versioned tool contracts. Attendees will see how to detect contract drift, validate scope and side effects, require confirmation for sensitive actions, block unsupported tool calls, and preserve an audit trail explaining every decision.

1. Evidence graph for microservices: How to unify traces, logs, deployments, and ownership metadata into a queryable cognitive index that supports investigations beyond dashboards.

2. Two concrete methods: How to perform deploy regression diagnosis using incident signatures and trace root localization, and how to compute change impact blast radius using explainable dependency paths.

3. Safe GenAI for operations: How to enforce tool grounded outputs with evidence references, confidence calibration, and refusal behavior when data is incomplete, plus how to evaluate accuracy against baselines.