Enterprise digital transformation often promises a streamlined future, but the immediate reality is frequently a fractured web of best-of-breed software platforms. For global enterprises, managing cross-functional workflows across fragmented point solutions has become a primary bottleneck to operational velocity and compliance.
In our recent webinar, Ben Shanahan, Leader of Procurement Transformation at Kyndryl, joined Spend Matters Chief Research Officer Pierre Mitchell and ORO Labs’ Chris Vessey to detail how a world-class technology services provider built a scalable framework for agentic AI by fundamentally rethinking enterprise procurement orchestration.
When Kyndryl spun off from IBM, they inherited an incredibly complex technological reality: 1,800 legacy applications operating globally across 60 countries. Over an intensive modernization window, Kyndryl successfully consolidated that footprint down to roughly 300 core enterprise platforms.
While they deployed market leaders including SAP Ariba, Workday, ServiceNow, and OneTrust, they hit a wall common to many Global 2000 organizations. These systems operate exceptionally well within their individual silos, but they rarely communicate smoothly with one another.
The resulting gaps led to "swivel-chair" data entry, manual patches, and growing frustration from internal stakeholders. Requesters viewed procurement as a non-transparent "black box" where visibility vanished and cycles stalled. Given that Kyndryl is a premier backend infrastructure provider that enables enterprise AI globally, they were arguably the most technically demanding customer ORO Labs could partner with. They knew that overlaying AI tools onto a fragmented environment without a solid architectural foundation wouldn't solve the friction, but would amplify it.
"If you have fragmented processes and fragmented systems, and you throw an AI agent into the mix, it doesn't solve the fragmentation—it expands on it," Shanahan observed. "For us, it was very thoughtful and sequential... Applying orchestration has put us in a really good spot now to just apply change, embed change in the workflow through a plug-and-play basis, and continue to grow."
To move away from rigid, hard-coded SaaS workflows that create brittle dependencies, Kyndryl established an integrated risk framework, mapped its end-to-end purchasing lifecycle, and deployed ORO Labs as its unified orchestration layer.
Shanahan detailed the four guiding principles that informed their deployment, showing how ORO turned theoretical procurement strategy into operational reality:
Kyndryl required a single front door for all third-party purchasing requirements. Whether a user is initiating a complex, multi-million dollar direct supplier project or an everyday purchase, every request originates from a singular intake management environment. This immediately removed the guesswork for employees trying to navigate various back-end platforms.
A major friction point in enterprise purchasing is redundant data collection. Individual process owners frequently build standalone intake forms that replicate the same foundational questions. By utilizing ORO, Kyndryl captures rich context at the point of request and maps it across systems. For example, by integrating with Workday, a user simply inputs a pre-approved project ID, and ORO automatically populates the corresponding procurement data fields—preventing repetitive manual entries.
With up to 50 conditional processes governing a transaction based on risk, geography, and cost, manual compliance is highly prone to error. To truly see how a decentralized environment can be secured globally, you can read about how global enterprises automate risk & compliance workflows to protect operations.
"We didn't want compliance to be guesswork," Shanahan noted. "The orchestration really became like bowling with the bumper lanes up. It prevents you from falling into the gutter; it just keeps you headed down the right straight path."
To eliminate the legacy "black box" stigma, Kyndryl exposed the entire workflow to the end user. Every necessary check, the specific internal owner, and measurable SLAs are visible in real time. This transparency creates dual accountability: business units can track their orders, and procurement teams gain clear visibility into process cycle times.
A composable procurement architecture separates the user engagement layer from underlying backend transactional systems. By utilizing an independent logical data layer, global enterprises can connect, modify, and optimize multi-system workflows without forcing expensive rip-and-replace technology projects.
This methodology aligns with current research on high-performing procurement ecosystems. Spend Matters’ Pierre Mitchell emphasized that true orchestration requires separating your orchestration platform from underlying point solutions.
According to Mitchell, legacy procurement applications often rely on rigid data structures that force companies into walled gardens. Modern procurement requires a composite application approach—an independent logical data layer that unifies enterprise data and processes without requiring teams to rip-and-replace their current technology infrastructure. Furthermore, Mitchell points out that as enterprise applications introduce standalone AI capabilities, managing a fragmented network of disconnected chatbots creates severe governance risks. True innovation requires combining predictive generative AI with deterministic, trusted business logic—a concept Mitchell highlights as key to building robust, automated workflows. [To discover how Mitchell and his team evaluate this structural shift, see how ORO exceeds leader benchmarks in key intake & orchestration criteria in the 2026 Hackett SolutionMap.]
The results of Kyndryl's orchestration strategy highlight the speed and scalability of the platform. Developed over a six-month deployment window, the system went live in mid-April. Within its first few weeks of operation, the platform achieved remarkable engagement:
With an orchestrated data layer firmly established, Kyndryl is uniquely positioned to maximize agentic AI. Rather than introducing independent AI tools that create further siloed complexity, Kyndryl can easily plug and play autonomous capabilities—such as automated RFP creation and intelligent historical pricing advisors—directly into their governed ORO workflows.
You can explore the full spectrum of these autonomous features by visiting our deep dive into Agentic AI for Procurement.
For enterprise procurement teams navigating complex software ecosystems, Kyndryl’s success offers a clear blueprint: establish a unified orchestration layer first, simplify your data foundations, and build a scalable path toward automated, intelligent procurement.
Watch the full webinar here.
Ready to see how orchestration can transform your business? Request a Demo of ORO today.
Expert contributors (webinar participants):