Five-week AI proof of concept routes customer requirement updates to their owners, then goes to production
A large automotive and industrial components manufacturer. Outcome verified
The problem
The quality organization of a large automotive and industrial components manufacturer must track a constant stream of customer-specific requirement updates, spread across many documents and formats, and decide which process owner has to act on each. Finding what changed and who cares meant slow manual reading. Ownership rules lived in people's heads, and the company wanted proof of value before a broad AI deployment.
Why it mattered
A missed or late interpretation risks audit findings under the automotive quality standard and non-conformance with major customers, and the document volume kept growing.
What we did
Vizius scoped a tightly bounded proof of concept: a portable tool that reads newly uploaded requirement documents, summarizes each change with a link to its source, and assigns it to primary and secondary process owners. Building in Vizius's own environment allowed weekly iteration.
What changed
The proof of concept was delivered and accepted on schedule in about five weeks, giving process owners an updates report organized by process. Three months later the client contracted to turn it into a monitored production application with review effort, routing accuracy and adoption as its measures.
- Manufacturing
- Large company
- IATF 16949