How do you provide subject-matter support for an ERP implementation in life sciences, so that GMP-critical processes are mapped correctly rather than becoming a risk only at go-live?

An ERP implementation in life sciences follows a fixed sequence, and the bottleneck moves with you if you break it. First, the subject-matter requirements must be precise enough that every GMP-relevant function has a testable target state: status changes from quarantined to released, quarantine logic, batch release as a prerequisite for QP certification under EU GMP Annex 16. The user acceptance testing builds on this target state, and only a UAT that also tests the negative cases shows whether the configuration under EU GMP Annex 11 actually works. Only after that does the performance qualification carry weight within validation, because it builds on the scenarios already confirmed from the subject-matter side instead of repeating them. If you start any of these stages too late, you push the untested remainder into go-live.

This is precisely the sequence we focus on. Data integrity is defined from the subject-matter side before IT configures: segregation of duties, the authorisation concept and the audit trail under FDA 21 CFR Part 11 are a specification from the subject-matter side, not a matter of interpretation for the administrators. This puts the effort where corrections are cheap, in the requirements and in the UAT, and not in the hypercare phase, when production is already running and every workaround outside the validated system creates a new risk.