Captur - intelligent extraction

Describe the fields. Captur finds them.

Captur turns paper, PDFs and email attachments into structured data inside Foldr. Tell Captur the fields you need; it locates them on every document, no template required for most jobs. For high-volume identical layouts, build a template and pin the zones. MaSH pushes the result to the share, system, or person that needs it.

purchase-order.pdf Extracted
No template Native PDF
Fields
  • Supplier Hopper Distribution
  • Order no PO-1188
  • Date 23 Jan 2026
  • Total £4,820.00
Watch

Watch a specification filed on your own lists

A specification is read and filed against your own coded list, and the one value Captur won't guess waits for a person's choice.

The full walkthrough: suppliers' specifications in several languages filed on your lists, laminate layers, the values left for a person, and how an admin sets the lists up.

1:23 · narrated · captions included · All videos

How it works

From the scanner to wherever the data needs to be.

Captur isn’t a separate inbox or a parallel database. The capture, the fields, and the file all live in Foldr, inheriting the same permissions as the share they land in.

1 · Capture

From the scanner, your inbox, or a folder Foldr crawls

Drop a PDF in, scan to a share, or send it to your Foldr inbox. Point a multifunction device at a share Foldr crawls. Captur picks up the file the same way regardless of where it came from.

2 · Extract

Captur finds them, whatever the layout.

Tell Captur the fields you need, supplier, order number, total, and it locates them on every document, regardless of layout, no template required. Build a template when a high-volume job warrants one: a zone can read a tick-box, a barcode, or (for handwriting and poor scans) a fresh read of just that part of the page. Line items come back as a list in a multi-value field, one row per item, and a total split across a page break still reads as one number. Text-based PDFs are read directly from their embedded text, with no OCR overhead. Scans and image-only PDFs go through OCR: Foldr’s own OCR service on Foldr Cloud, your own cloud OCR account if you’d rather, or Tesseract on a self-hosted appliance. Once transcription is on for a share, Captur reads a recording’s transcript as pages too.

Captur uses a language model to read each document. On Foldr Cloud that runs on an EU-hosted route, Enterprise customers can bring their own provider, and on an appliance you can point it at a model that runs on your own hardware, so nothing leaves. The Trust Centre has the details, and so does the plan: Captur is part of Enterprise on Foldr Cloud and a licence key on an appliance.

3 · Route

Where the data ends up is up to you

Filed by supplier and date. Pushed into a finance system over MaSH. Emailed back to the submitter for review. Same Captur template, different routes per share, all configurable without code.

+ Verify

Humans where it matters

Every value lands as an editable custom field on the file itself, not in a hidden inbox. Anyone with access can open the document, see what Captur read, and correct a field in place. Once a person has corrected a value Captur leaves it alone, and when it read something two ways it asks you which was right instead of picking.

With Grace on top

Ask Grace about everything Captur captured.

Once Captur lands data in custom fields, Grace can read it. “Show me purchase orders over 5,000 from last quarter.” “Which delivery notes are missing a signed receipt?” Grace stays inside the user’s permissions, so a buyer only sees their own POs.

Curious whether Foldr would fit?

Try it free for 30 days, no credit card. Or get in touch about self-hosting.