Case 02 / P2P · AI controls · Supplier management
AI supplier controls that released €2M of working capital
AI applied to invoice duplicates, invoice quality errors, supplier statement reconciliation and supplier queries, on top of a supplier lifecycle portal.
01 The problem
Duplicate payments, invoice quality errors and unreconciled supplier statements are expensive and hard to catch by hand. Supplier queries add a steady load of manual work on top, and supplier data problems feed every step that follows.
02 Who it was for
Accounts payable and supplier teams, the controllers accountable for payment accuracy, and treasury, for whom every prevented error is cash that stays in the business.
03 My role
Global process ownership for Procure to Pay: making the case, setting the control design, and tying each AI module to a measured FTE or cash outcome.
04 The process
The foundation was a supplier lifecycle portal for vendor master data creation. AI controls then covered four areas: invoice duplicates, invoice quality errors, supplier statement reconciliation and supplier query handling.
05 The numbers
06 Key decisions
- a.Fix supplier data first, since every later control depends on it
- b.Tie each AI module to a measured FTE or cash outcome
- c.Count working capital separately from cost savings
The result
€2.05M of yearly working capital improvement, 21 FTE of capacity released, ≈29,000 hours a year, €927k in yearly savings and €369k in cost avoidance.
07 What I learned
AI earns its place in the P&L when it is pointed at a specific, measurable leak. Each control here has its own number attached.