AI success in procurement hinges less on the sophistication of the model than on the quality of the data it can access. That is the central message running through the article series: without clean, structured and connected information, even the most capable systems will struggle to produce dependable results.
Procurement teams are still wrestling with fragmented platforms, inconsistent formats and incomplete transaction histories. In practice, that means data is often trapped ...
Continue Reading This Article
Enjoy this article as well as all of our content, including reports, news, tips and more.
By registering or signing into your SRM Today account, you agree to SRM Today's Terms of Use and consent to the processing of your personal information as described in our Privacy Policy.
in separate tools across sourcing, contracting, supplier management, intake and accounts payable, leaving AI with an unreliable view of the business. As TechTarget has reported in related coverage, ERP fragmentation can weaken governance, damage confidence in reporting and make modernisation efforts harder to sustain over time.
During the webcast “From Systems of Engagement to Systems of Action: The Agentic CPO”, Andrew Bartolini of Ardent Partners and Saquid Jawed of Zycus argued that procurement data needs to go far beyond basic transactional records. It must also capture the context around those transactions: the supplier relationships, approval routes, category strategies and contract terms that explain how a decision was reached. That richer layer of information gives AI the ability not only to register what happened, but to infer why it happened.
A key idea in that discussion was the importance of decision traces. These are the records of the logic behind procurement choices, including supplier selection, escalation paths and approval rationales. Preserving that evidence is increasingly important because it gives AI systems a framework for learning how human teams operate and how future decisions might be improved. In effect, it turns procurement history into a usable intelligence asset.
The broader case for AI-ready data has been echoed elsewhere. At Gartner’s 2026 Data and Analytics Summit, experts stressed that organisations need data that is prepared, accessible and properly governed before AI can be expected to deliver value. Similar points have been made in other industry commentary, which notes that data quality, integration and governance are not optional preliminaries but prerequisites for meaningful AI deployment.
Governance, however, is not simply about slowing automation down. In an agentic environment, it becomes a way of defining boundaries: what systems can decide for themselves, where human oversight remains essential and which thresholds trigger intervention. Rather than replacing control, this model redistributes it, allowing procurement teams to set the rules while software executes within them.
That matters because the biggest obstacle to progress is often not the AI itself but the architecture surrounding it. Multiple disconnected systems create blind spots, duplicate records and conflicting versions of the truth. GEP has warned that fragmented procurement environments can undermine board confidence through missed renewals, shadow spend and weakened negotiating leverage. In multi-ERP enterprises, as Velocious has noted, supplier records and approval workflows can become so dispersed that no single team has a complete view.
The direction of travel, then, is towards a more integrated operating model in which procurement platforms connect the entire source-to-pay cycle. NTT DATA has argued that agentic AI depends on high-quality, context-rich and timely data, alongside a clear understanding of how decisions are made. That same logic applies in procurement: if data flows cleanly across systems, AI can optimise processes end to end rather than improve only isolated tasks.
Bartolini and Jawed’s vision points to a future in which procurement is shaped by autonomous sourcing, intelligent intake, live supplier management and dynamic contract enforcement. In that model, procurement professionals spend less time pushing paper and more time orchestrating outcomes. But the article series makes clear that such a shift will only work if organisations first get the fundamentals right: data quality, system integration, governance design and a disciplined way of measuring business value.
Automation alone will not make procurement intelligent. The real prize lies in turning scattered information into a continuous decision-making engine that is both autonomous and accountable.
Source: Noah Wire Services
