For finance, operations and IT

AI document data extraction: the model reads the document, a person decides on the flagged cases.

Invoices, receipts and incoming documents are read with Document Intelligence, and a model in Microsoft Foundry extracts the fields and writes them to your ERP. It runs in your own Azure, in an EU region.

What we build

The same pattern in every project, the document read, the fields validated, uncertain cases flagged for a person.

  • Documents read and extracted

    Document Intelligence reads the text, and a model extracts the supplier, VAT number, lines, VAT and totals.

  • Confidence threshold

    Results above the threshold are saved automatically. Those below it are not posted: they are flagged for a person to decide.

  • Connected to your ERP

    The result is written to your ERP, through an integration, or through a robot where there is no API.

How we build the extraction

Real documents and the confidence threshold first, then the volume.

  1. 01

    The documents

    Which documents, which fields, who decides.

    • Teams
  2. 02

    Model and threshold

    The model that fits the task and the cost, the threshold below which a person decides.

    • Azure
  3. 03

    The integration

    With your ERP or SharePoint.

    • ERP
    • SharePoint

Show us the documents someone still types in by hand.

In 30 minutes we go through the process and agree what to prepare before the project.

Denis Stadler
The first call is with Denis Stadler.Founder · AI Architect & Fractional CTO