Demo on fictional data
Receipt and invoice extraction
This is a demo on fictional data that shows receipts turned into fields (date, vendor, total, tax, category) with a confidence score per field, and a needs-review queue for anything the rules are unsure of.
Requests since you opened this page: 0
Works offline: not installed
The browser policy on this page forbids network requests (connect-src 'none'). The count is here so you can see it hold. How the demos work.
The receipts
Ten made-up receipts for Sample Landscaping Co., as plain text. The extractor below reads them with regular expressions and a vendor lookup table, which stand in for the model step production would use.
Extract and review
Any field whose confidence is under the threshold sends the row to review. The confidence is the rule's own measure: a labelled total scores higher than the largest dollar amount on the page.
| Receipt | Vendor | Date | Total | Tax | Category | Status | Actions |
|---|
Approval log
| Time | Receipt | Decision | Approver |
|---|
What this shows, and what it does not
What this shows
- A rules-based extractor reading text receipts into fields, each with the rule's own confidence score.
- Rows under the threshold are flagged for review instead of being guessed.
- Every approve, fix or reject decision is logged with an approver.
What it does not show
- No model reads these receipts. Regular expressions and a lookup table stand in for the model step production would use.
- The confidence score is the rule's own measure, not a model probability.
- No number here is a result a client achieved.
How it would run for you
This is the shape of a Prototype Sprint: the agent drafts, a named person approves, the log stays with you.
Answers the question on the Texas or Taxes tile in the Fridays campaign.
Related services: Prototype Sprint, Friday Sprint.