AI is becoming part of nearly every logistics technology stack, but investment alone doesn’t guarantee impact. Val Marchevsky, CTO of Uber Freight, will discuss how logistics AI has moved beyond pilots and surface-level use cases—applying it in ways that improve decision-making, accelerate tailored solutions, and create measurable value across complex freight operations.

Most supply chain operations follow a predictable pattern in which 80% of work is repetitive and rules-based, while 20% requires judgment, context, and decision-making. AI is exceptionally good at the 80% of repeatable task-level work, but that’s not where operations succeed or fail. In this session, David Lee, Sales Engineer VP, examines the gap between where AI performs best and where operational risk lives. Using real-world logistics scenarios, he will explore how organizations are rethinking the orchestration of work between humans and AI to improve efficiency, protect service levels, manage exceptions, and maintain control in volatile environments. Attendees will gain a clear understanding of how to deploy AI where it creates the most leverage, and where human expertise remains invaluable to supply chain operations.

Every supply chain department has workflows that should have been automated by now. The disconnect between automation ambitions and the manual work teams still handle is where many organizations get stuck. You’ll learn how Fortune 500 companies streamlined quote management workflows, including a proven adoption framework based on real-world deployments and a free tool to analyze your own shared inboxes and identify where to start.

As organizations race to adopt AI, financial systems demand a different standard: predictable outcomes over clever answers. Tony Urban of Cass discusses why determinism remains critical for auditability, trust, and scale, and where AI can create value without introducing risk. In financial processes, organizations should be clever where they build AI—and deterministic where they run it.

A fake Pentagon attack image erased billions in market value in minutes. A false tariff rumor sent $5 trillion in trades through the markets in half an hour. AI untethered from reality is no solution at all, on Wall Street or in freight. GenLogs' Isaac Winnes explains why automation is only as good as its inputs, and how ground truth keeps freight AI anchored to reality.

The dark warehouse is a keynote fantasy with bad unit economics. George makes one claim in ten minutes: the facilities pulling ahead right now pair heavy automation with people who own the judgment calls. He proves it with two deployments he ran personally, one from his founding years at Nimble Robotics and one from a live Gather AI site, showing exactly where the human sits in each loop and why removing them would have made the operation slower and more expensive. He leaves the stage with a simple test any facility leader can apply to their own automation roadmap.