The receipt is the expense claim.

Your team photographs the receipt at the table and stops thinking about it. Marlow flattens it, reads it and codes it, so month end arrives with nothing outstanding and nothing guessed.

Try the capture demo Talk to sales

11 days

Faster month-end close

96%

Receipts captured within 24 hours

Spreadsheets in the approval chain

Built for the people who chase the receipts

Capture that survives real life

Creased, folded, photographed badly on a dark table. We straighten it, correct the colour and read it anyway, so nobody is asked to take it again.

Figures you can check

Every value we read is shown on the receipt where we found it. Your reviewer confirms a total at a glance instead of opening the original and hunting.

Ready for your ledger

Coded to your chart of accounts with the VAT split out, and posted to Xero, Sage or NetSuite without anyone rekeying a line.

Three steps, one of which is yours

Photograph it

Any angle, any light, straight from the phone. That is the entire task your team is asked to do.

We clean and read it

The receipt is found in the photo, flattened, colour corrected and read, in that order, because each step makes the next more accurate.

Finance approves

The claim arrives coded, with every figure highlighted on the image it came from. Approve or query in one click.

“The thing that changed was not the scanning. It was being able to see the number highlighted on the receipt. Our approvals went from a morning to about twenty minutes.”

Financial Controller, 240-person engineering consultancy

Photograph a receipt and watch it work

Use the worst receipt you can find. Crumpled, angled, badly lit. See it flattened, corrected, read, and every figure marked where it was found.

Try the capture demo

Receipts are flattened, colour corrected and read by the Filestack Processing and Intelligence APIs. Bounding boxes come straight from the OCR response.

What Filestack does here

Finds the receipt in the photograph and flattens it, corrects the lighting, then reads every figure and remembers where on the page it found each one.

View the source