All case studies. Browse by sector

DataCactus dataCactus
Case studies How we work
Talk to us

Case studies · Charities & care

Referral automation for a special-education and care provider

Referral and consultation bundles arriving as PDFs and Word files are now read, classified and turned into a drafted internal assessment, ready for human review and approval.

Client
A special-education and care provider
Built with
Power Automate, Azure AI Foundry (GPT and Mistral models for OCR and classification), SharePoint, Dataverse

The problem

Referrals and consultations arrived by email as bundles of PDFs and Word documents, often including scanned pages. Each one had to be opened, read, classified and logged by a person before anyone could assess it, and the internal assessment was then written from scratch against a deadline.

What we built

  1. 1Automatic pick-up of incoming documents from the mailbox and SharePoint
  2. 2Every page read, including scans and photographs
  3. 3AI classification of document types and extraction of the person, the referrer, the needs described and the dates
  4. 4The provider’s own internal assessment documents filled in from the extraction, with the key referral points surfaced for the reviewer
  5. 5An approval workflow in which a person reviews, corrects and approves before anything is sent or recorded

What changed

  • —The team starts from a completed internal pack rather than a mailbox
  • —Every referral has a consistent record of what was received, what was extracted and who approved it

This project became a ready-made offer you can start from.

See the offer

More in Charities & care

HR analytics pipelines for a national charity A national health charity Corporate services source of truth for a social-impact charity A national social-impact charity Sector page Charities & care Referrals, reporting and back office for providers and charities