
November 23–27, 2026
Deloitte Consulting
Most talks about agentic AI start with a clean slate — new tools, fresh data, no history to work around. Large companies don't get that luxury. SAP is the software that runs the back office for thousands of big organizations: finance, supply chain, HR, all of it. It's been built up and customized over decades, which means it comes with a lot of complexity baked in. SAP Joule is SAP's answer to the AI moment — an assistant meant to not just answer questions, but actually take action inside that system. This talk is about what happens when you try to bring an AI agent into a system like that. Before an agent can be trusted to act, a few things have to be true. The underlying data has to be organized well enough for the AI to make sense of it. Someone has to decide exactly what the AI is and isn't allowed to touch. And there's a real difference between an AI that can answer a question and one that's safe to let loose on actual business processes. This isn't a product demo. It's a plain look at where the promise of AI agents runs into the reality of how large enterprises actually work — and what it takes to close that gap. No SAP background needed. By the end, you'll understand the core challenges of putting an AI agent to work inside any large, established enterprise system, using SAP Joule as the real-world example.
See every session across the AI Agents Conference November 2026 line-up.
View Agenda