Shane Wiggins

AI is moving from giving recommendations to taking action. Agents now query systems, call tools & APIs, and trigger other agents on their own. Building and deploying them has gotten much easier, and many are heading toward a point where agents outnumber the people managing them. Traditional governance wasn't built for that kind of scale or autonomy. Knowing which agents exist isn't enough. Enterprises also need to see how agents interact with data, applications, APIs, models, and each other and what that activity costs. Every agent action consumes tokens and compute, and that spend can grow quickly and invisibly across a large agent fleet. Without visibility into usage and cost alongside behavior, organizations can lose control of risk and budget at the same time. Agent trust solutions close that gap. Instead of just documenting what agents are supposed to do, observability shows what they actually did: what data they touched, whether it met quality standards, which policies they followed, and how much it cost to run. That evidence is what lets enterprises attest to agent behavior with confidence not just assume it. Together, these signals including data quality, policy adherence, execution history, and cost per action create a measurable model of trust. That's what allows enterprises to scale AI agents safely, keep spend under control, and stay accountable as autonomous systems take on more of the work. This session explores what enterprises need to observe, how to monitor…

As organizations scale data and AI initiatives on Databricks, speed and trust must advance together. This session explores how modern enterprises can operationalize trusted data and AI by combining security principles with actionable governance guardrails. Learn how the Databricks Data and AI Security Framework can be extended with policy‑driven controls that ensure consistent access, use, and governance of data across analytics and AI workflows. Join us for this discussion and live demo to see how guardrails can be applied across the AI lifecycle to help organizations manage risk, respect data‑use requirements, and demonstrate compliance without disrupting innovation. Attendees will gain practical insight into turning governance from static policies into enforceable, auditable controls that scale with Databricks — enabling confident, responsible data and AI initiatives at enterprise scale.