SAP AI Agents in Production: Governance, Joule Studio and Accountability
Getting an AI agent to work in a demo is one thing. Letting it act inside a live business process is another. Before you scale, you need answers to harder questions: Who owns it? What can it access? How do you know it's doing the right thing? And who is accountable when it isn't?
Across three days, learn what it really takes to move from experimentation to production. Hear how Queensland's Department of Transport and Main Roads is approaching threat modelling and accountability, what Fonterra learned from putting Joule in front of 100 users, and how SA Power Networks is governing AI across its SAP landscape. Go deeper with SAP on Joule Studio, permissions, lifecycle and evaluation, so you leave with a more practical framework for deploying agents safely, governing them effectively and proving whether they are delivering value.
Designing Your Enterprise AI Strategy with SAP
DEEP DIVE · 90 MINSteven Chua and Swamy Bathina, SAP
Ninety minutes on the architecture that lets agents scale past pilots: Business AI, Joule, AI agents, clean core and BTP working together.
Paul Hesford (Queensland Department of Transport and Main Roads) and Phil Burgess (SAP)
How TMR used process-level threat modelling, clear accountabilities and an incident exercise to prepare for AI inside critical SAP processes.
SAPinsider APAC 2026 · Innovation Vanguard finalistSAP Best Tech Awards 2026 · Intelligent ERP finalistNainesh Golchha, Fonterra
What Fonterra learned from rolling Joule out to more than 100 users, including where it delivered value, where it didn't and what was needed to sustain adoption.
Travis Smith, SA Power Networks
How SA Power Networks is creating one architecture and governance model for AI across SuccessFactors, S/4HANA, Concur and BTP.
Rachel Hunter, Thorsten Freitag and Murali Shanmugham, SAP
An end-to-end view of SAP's AI stack, from embedded AI and Joule to orchestrated agents and bespoke solutions grounded in business data.
Shaun Bradridge, Independent
How to decide whether a customer-facing process needs AI, conventional automation or straightforward application logic.
Katan Patel and Jorg Thuijls, Independent
How multimodal AI and n8n were connected to Fiori to turn warranty attachments into structured business insights.
SAP TechEd curriculum
How Joule Studio brings agents, automations, integrations, lifecycle management and governance into one build environment.
SAP TechEd curriculum
The operations session. Usually the one that's missing.
SAP TechEd curriculum
How scoped APIs, runtime policies, human approvals and observability govern agents acting across SAP and third-party systems.
SAP TechEd curriculum
What to do about the agents your teams have built without telling you.
SAP TechEd curriculum
How to establish ownership, runtime controls and central visibility as agents, models and MCP servers spread across the business.
SAP TechEd curriculum
How to know whether it is working.
SAP TechEd curriculum
How agent identities and managed access controls enable secure communication across SAP landscapes.
SAP TechEd curriculum
How Joule spans a multi-system landscape.
SAP TechEd curriculum
How domain models and business semantics give agents the context to understand processes, navigate data and make informed decisions.
SAP TechEd curriculum
What SAP means by autonomous, in product terms.
A practical production model covering agent ownership, approvals, evaluation, monitoring and human oversight, plus a clearer way to decide which processes genuinely need an agent.
AI lead · Enterprise architect · Process owner · Risk or internal audit
Pre-conference 17 Nov · Main conference 18–19 Nov · ICC Sydney · Agenda and timings subject to change