Master data, outline agreements and release strategies harden into the system at go-live. This page maps the decision rights that must be owned before the architecture question closes — for regulated procurement in pharma, medtech and animal health.
01
The go-live date is a governance deadline.
Situation: the migration plan is on track; the ownership plan is not. Risk: whatever is unowned at go-live — master data, release strategies, contract control — hardens into the system for years. Intervention: a bounded review of decision rights before the architecture question closes. Boundary: decision support — not implementation, not certification.
02
Master data needs an owner, not a committee.
Situation: vendor master and purchase-contract data sit between procurement, finance and IT. Risk: every unowned field becomes a dispute at the worst moment — during an audit or a supply break. Intervention: ownership logic per data domain, drawn from practice as data domain owner for SAP purchase contracts. Boundary: employment-context experience, not a current client reference.
03
Outline agreements under control.
The published S/4HANA outline agreement control case documents the pattern: contract control designed into the system instead of patched after go-live — USD 450K reported cost avoidance, Class B in the evidence registry.
04
Controls that survive an audit.
12 P2P SoX controls developed and implemented in a carved-out, regulated business (Elanco).
Go-live of a standalone cloud system with hypercare leadership.
Near hyper-automation of invoice processing without additional software inside 12 months.
05
Governance at group scale.
Merck Group, Darmstadt (2025–2026): governance architecture for the end-to-end Source-to-Pay process across the whole group.
Blueprint for the GPO S2P role archetype for Group Procurement.
The same discipline, applied where S/4HANA decisions live.
06
The honest boundary.
These are employment and assignment contexts — publicly verifiable career facts, no current client relationships implied. This page supports the decision before the system hardens; it does not implement, validate or certify.