Where Vrocure uses AI, that those results can be wrong, and how to report a bad one. The model provider is Anthropic.
Some screens in Vrocure draft, explain, or check work with AI. The provider is Anthropic. Treat the result as a draft. It can be wrong.
Vro, the assistant inside the signed-in portal.
Dr. Review findings on a drawing.
Bid copilot, when you compare quotes.
Discovery match explanations.
The spend narrative on analytics.
Order intake when you fill a new order from a PDF, an email, or a written description.
Process checklists generated for a job.
The machining guide.
The help assistant on this site and in the portal.
A spreadsheet or CSV import maps columns you choose. That step does not use AI.
Dr. Review is an AI-assisted check, not an engineering sign-off. Verify the findings against the drawing and the standards that apply before you release the part.
An AI result can miss a note, invent a detail, or explain a match poorly. Read it before you rely on it.
The provider is Anthropic. By default, Anthropic does not use inputs or outputs from its commercial API to train models. The current wording, including when that default does not apply, is on Anthropic’s privacy article.
If a result is wrong, email support@vrocure.com. Say which screen you were on and what looked wrong. That address is the support inbox in the product.