AI System for Reviewing the Impact of Quality Document Revisions

The problem

When revised documents, relationships, processes and training inventories are held in separate sources, their fields must be matched by hand before a list of affected documents and responsible owners can be prepared.

The quality systems manager must compare every suggestion in the list of affected documents and responsible owners with its source record; missing or conflicting evidence must not replace the final decision.

If similar wording that applies to a different site is not set aside for review, a critical linked record could be missed; the record must not proceed without a human decision.

What we set out to improve

• Time taken for a draft list of affected documents and responsible owners, together with its supporting evidence, to reach the quality systems manager for review
• How records set aside for human review because similar wording applies to a different site are resolved
• Number of suggestions corrected after identifying a risk that a critical linked record could be missed

How the system works

1. 01 Retrieve the revised document, relationships, processes and training inventory from permitted sources using the record identifier
2. 02 Link source, date and version information to the shared work record
3. 03 Validate fields and units
4. 04 Prepare the list of affected documents and responsible owners with its supporting evidence
5. 05 Check for an exception: similar wording applies to a different site
6. 06 Have the quality systems manager accept, revise or reject the suggestion
7. 07 Record the approved decision and its source in the processing log

Methods we used

• A source-identified common data schema for the revised document, relationships, processes and training inventory
• A decision structure that separates deterministic rules from AI suggestions
• A review screen showing the list of affected documents and responsible owners alongside the underlying source evidence
• A mandatory human review queue when similar wording applies to a different site
• An event identifier that prevents duplicate processing

Where people stay involved

Data and security

Only the fields required for this decision should be processed from the revised document, relationships, processes and training inventory. Access to the source systems must be authorised separately from the quality systems manager role, and personal data, commercial information and technical secrets must not be passed into the model context unnecessarily. Reading, suggestions and human decisions must be logged separately.

Who this suits

Frequently asked questions