What We Learned at Coupa AI DevCon 2026
A recap from the floor of Coupa AI DevCon 2026: what's genuinely coming for AI in spend management, and the concrete steps customers should take now to be ready for it.
We spent Coupa AI DevCon 2026 in the sessions and at the booth, and came away with a clearer picture of where AI in spend management is actually heading — past the demos and into what teams will run day to day.
What's actually coming
The center of gravity is shifting from AI that summarizes to AI that acts. Agents that triage supplier risk, draft and reconcile invoices, and route approval exceptions are moving from roadmap to reality. The theme across every serious session was the same: the value is not the model, it is the connective tissue — clean data, well-defined actions, and guardrails that let an agent operate inside a real approval chain without going off the rails.
What customers should do to prepare
The organizations that will benefit first are not the ones with the fanciest models — they are the ones whose data is ready. Three moves matter now:
1. Fix your master data. Duplicate suppliers and inconsistent categorization are the number-one reason AI agents produce wrong answers confidently. 2. Define your actions and guardrails. Decide which decisions an agent may take autonomously and which require a human, before you turn anything on. 3. Close your integration gaps. An agent is only as useful as the systems it can reach; disconnected ERPs and portals cap what any AI can do.
The teams that treat this year as data-readiness groundwork will be the ones running production AI next year — not still evaluating it.
Questions about your systems?
Talk to an expert
