THE PROCUREMENT INTERVIEW
Yuan says this is also why older software has struggled to keep up. The architecture was never designed around orchestration and in his view that limits how well it can support the business today.
Agentic workflows in action The founders see agentic AI as the next step in that evolution. In their view, AI should handle repetitive, low-value tasks or circumstances where data needs to be pulled and analysed across many documents and systems, while humans focus on judgement, accountability and higher-order decisions.
The impact can already be seen in live customer deployments. ORO is working with pharmaceutical companies including Pfizer and Novartis on a PR to PO review process that used to be manual.“ In 70 % to 80 % of the work, AI is doing it,” Sudhir explains.
That matters because faster review cycles can speed up the business user’ s ability to work with suppliers, while also improving compliance. Where humans are more likely to miss details, AI can identify issues earlier in the process. One of the company’ s clearest examples comes from an oil and gas customer, where a supplier proposal can contain multiple line items and safetyrelated checks. ORO’ s platform can now read the proposal, work out the correct workflow and produce a purchase order in minutes, where once the process took hours. The overall result is a major reduction in manual work.
procurementmag. com 31