Ten Inspires in that you develop a feel for the difference between a company “performing” confidence and a company that actually has it.
Recently, Inspire felt like a conference in transition. The energy was there. The production was polished. But the strategy underneath carried the weight of a company still reconciling what it had been with what it was becoming under new ownership and new leadership.
Inspire 2026 felt different. Not louder or more expensive, just more settled. This year felt like the “new” Coupa, not the one finding its footing. The strategy had tight focus and coherence, and above all, ambition. Whether execution matches ambition is a separate question. But the vision, at least, has arrived.
Building on a $10 Trillion Data Asset
The foundation of Coupa’s strategy is a claim that deserves to be taken seriously.
According to CEO Leagh Turner, more than $10 trillion in spend has flowed through the platform over two decades. Every purchase order, invoice, contract, supplier interaction, and payment contributes to what Turner repeatedly framed as one of Coupa’s most significant competitive assets.
That data foundation is now being operationalized with the bold announcement of a proprietary Domain-Specific Language Model, or DSLM. Rather than relying exclusively on general-purpose large language models, Coupa is investing in AI that is trained specifically around its procurement, sourcing, supplier management, and spend data.
The strategic logic is compelling. General-purpose LLMs are becoming increasingly capable and increasingly accessible. If every provider is building on the same foundational models, differentiation has to come from somewhere else. Coupa is betting it comes from proprietary data and the intelligence built on top of it.
It is an ambitious bet. Building and maintaining a domain-specific model requires data science infrastructure, model governance, evaluation frameworks, and continuous retraining capabilities. These are not things ProcureTech providers have historically had to develop. Announcing a DSLM and demonstrating that it consistently outperforms general-purpose models on procurement-specific tasks are meaningfully different achievements. That distance is worth watching.
Still, the underlying logic is difficult to dismiss. If the $10 trillion data asset is as differentiated as the company believes, the DSLM is the right vehicle for turning it into something competitors cannot easily replicate.
Coupa Compose
The headline product announcement at Inspire was Coupa Compose.
Designed as the company’s framework for agentic procurement, Compose brings together three components. Navi Agent Studio provides a workspace for building and managing agents. Smart Intake and Orchestration connects user intent with automated execution. Navi Connect extends workflows into external systems and partner ecosystems.
Coupa currently offers approximately 20 agents and plans to expand that portfolio significantly. The number of agents matters less than the architecture being built around it. Compose is an attempt to establish a common framework for deploying, governing, and scaling AI across the broader suite. The objective is not task automation in isolation. It is coordinated workflow execution across sourcing, procurement, accounts payable, and supplier management.
Throughout the conference, executives were consistent on this point. The future of procurement AI is not individual assistants operating independently, but rather, orchestrated systems of intelligence capable of supporting larger business processes. Compose is Coupa’s answer to that future.
Catalyst
One of the more thoughtful announcements at Inspire was not a product. Coupa introduced Catalyst, a customer transformation initiative combining forward-deployed engineers, business consultants, AI specialists, and rapid prototyping support to help organizations move from AI potential to operational reality.
Catalyst acknowledges something the ProcureTech market has been slow to admit. Technology alone rarely determines outcomes. Governance, data, process redesign, and organizational readiness often matter more than the software itself. The launch of this service shows Coupa’s alignment with its customers. The service makes particular sense given Coupa’s new commercial model (see below).
Rethinking Commercial Models
Coupa outlined that it is moving to outcome-based pricing and a commercial model tied to savings rather than traditional subscription metrics. The concept follows naturally from the AI strategy. If agents generate measurable improvements in efficiency, compliance, and savings, pricing should increasingly reflect those outcomes rather than seat counts or transaction volumes.
Coupa’s move to outcome-based pricing makes it one of the first to do so; however, the focus on savings as the metric to track and price (versus some type of transaction throughput or spend volume) will create implementation complexity. Questions around measurement, attribution, and consistency across different customer environments will take time to codify. Procurement organizations operate under widely different conditions and maturity levels. Determining what is attributable to software versus broader business factors is genuinely hard.
Still, the direction is the right one. The industry has talked about outcome-based software for years. A few major providers have moved meaningfully toward it. Coupa joins that short list.
Rossum and the Acquisition Question
Coupa highlighted its acquisition of Rossum, an AI-powered document intelligence provider, during the conference. Rossum fills a capability gap around document intelligence, although its long-term significance will depend less on the technology itself and more on how broadly Coupa extends it across the platform.
Shortly after Inspire concluded, Coupa announced the acquisition of Tonkean, a provider of intake and orchestration technology.
Both acquisitions align with the themes introduced through Compose. Both also surface a question that customers will be watching closely. As Coupa continues to expand through acquisition and internal development, platform cohesion is becoming increasingly important. Customers will be watching closely to see how quickly these capabilities evolve from adjacent products into a genuinely unified experience.
Customer Stories and Community
As always, customer participation was central to Inspire. Organizations, including Synchrony and Kimberly-Clark, shared their experiences with procurement transformation and AI adoption. Discussions centered on speed, operational efficiency, and the evolving role of procurement in enterprise strategy.
The conference also reinforced one of Coupa’s most durable strengths. Its customer loyalty. While there was a sub-theme at the conference regarding the network, Coupa has always had a large stable of happy and very vocal customers. Its community is a competitive asset, and Coupa knows it.
Looking Ahead
Inspire 2026 was one of the most coherent versions of Coupa’s conference in the ten years I have attended.
The different AI announcements (DSLM, Compose, Catalyst, and outcome-based pricing) fit together as parts of a larger vision rather than isolated announcements. The strategy has a logic to it. The community data could be a major differentiator. And the ambition is real.
The harder questions sit just beneath the surface. How the DSLM performs against general-purpose models in practice. Whether Compose delivers on its orchestration promise at scale and on what time frame. How the Rossum and Tonkean acquisitions integrate into a genuinely unified platform experience rather than an expanding collection of capabilities.
Those questions deserve a longer examination. We will take them up in Part 2.
RELATED RESOURCES
Coupa Inspire 2025: AAA Rating – Autonomous Trade, AI, and Acquisitions
Coupa Inspire 2024: From BSM to Margin Multipliers
