Extract structured data from invoices, receipts, delivery notes, and finance documents — then send it where your systems need it.
~3 min
for a batch of 100 invoices, typically
13
standard document types live
200 pages
free to try, no card
How a document comes in, gets read and checked, and goes out to your books on the rules you set.
Free for 90 days. One page uses one credit.
Is your team still capturing invoices by hand at month end? Every invoice typed by hand is another chance for a mistyped total.
So stop typing them, and forward the email or point us at the folder they land in. We read every field and every line item, and if the totals don't add up, it's flagged, so it can never auto-approve. Review each one, or switch on auto-approve for documents that pass every check. Then it goes wherever your books live, as a Xero bills file or through the API.
And if someone overrides a failed check, the activity log records who did it.
And that's DocReadi. We handle the documents, you handle everything else. Start your free trial now, test it with last month's invoices, and see the results for yourself.
Stop retyping supplier invoices. The numbers are already printed. One typed total can be wrong. You find out when the books don't balance.
Send them in. Email, upload, the API, or a watched folder. We read the fields and the line items. We check the maths. Not on the page? Left blank. Review every document. Or auto-approve the clean ones. Pick what each export sends, and where. Files, the API, webhooks, or a folder. You stay in control of every number. Every workspace kept separate.
Review by default. Documents wait for review unless you switch on auto-approve. Nothing invented. Absent values stay blank. Missing required fields are flagged. Checked arithmetic. Line items must add up to the subtotal, and subtotal plus VAT to the total. Isolated workspaces. Every workspace is isolated from every other at the database level.
We handle the documents. You handle everything else. Start your free trial. No card needed, at docreadi.com.
Upload on the web, use the API, or forward your supplier inbox to a DocReadi address. PDFs, JPG and PNG images (scans and phone photos included), and e-invoice XML. WhatsApp works too, once you connect your own Meta Business number.
We read the fields and the line items, and check them against what we know about the counterparty. A new bank account, a new address or a total that doesn't match the line items is flagged for review.
Your data goes out as JSON via the API, signed webhooks, CSV or Excel exports, or a Xero Bills CSV. You decide which.
Line items, totals, VAT, supplier bank account, payment terms.
What your customers received from you, line by line.
Refunds and adjustments, with the original invoice number when it's printed.
Line items, totals, expected delivery date and terms.
Items and quantities delivered, delivery date, order reference, and who signed for it.
Store, date, total, line items where present, VAT.
We record a supplier's bank details, VAT number and address. Once the supplier is confirmed, every new invoice is checked against that record. When a bank account or address changes, the warning names the field and the new value. Only an admin can clear a new-bank-account warning.
A supplier we haven't seen before starts as a candidate. After three consistent invoices it is confirmed automatically. You can change the number, or confirm suppliers by hand.
A structured reference with bearer-token auth and documented rate limits. Included on every plan, the free trial too.
Read the API quickstart →Signed POSTs to your endpoint on the events you pick, such as a document approved, rejected or failed. Failed deliveries are tried up to six times over about seven hours.
CSV and XLSX in one click. Filter by counterparty, date range, status. Scheduled exports to S3 or email.
Your team approves, rejects or edits flagged documents inside DocReadi. Your edits are what the API and every export return.
What we do. Read business documents. Extract structured fields. Learn counterparties over time. Detect drift in bank accounts and addresses. Hand the structured rows to your stack.
What we don't. Approval chains. AP automation. Reconciliation. Payment release. We feed the systems that already do those — Xero today, and any other system through webhooks and CSV / XLSX exports.
More answers in the help guide.
No. We don't train models on your documents. To read them, we send them to the AI and OCR providers listed on our sub-processor page, along with what each provider's terms say about training.
We tell you. The document is marked Failed, and its page shows which step failed and the error. You can reprocess it, upload a clearer copy, or contact support.
We host in the United States (Railway, us-west2), and traffic reaches us through Cloudflare's global network. Every workspace is kept apart at the database level. Saved bank accounts, connection passwords and secrets are encrypted in the database. Two-factor authentication is available on every account. For a specific POPIA or GDPR question, talk to us.
We score confidence per field, and a document scores as low as its weakest field. There isn't a single accuracy number that's honest. By default every document lands in your review queue. When you're ready, turn on auto-approve per document type and set the threshold. It's opt-in, not something we decide for you.
Xero, through a Xero Bills import file. For anything else, use webhooks plus CSV or Excel exports, and pick your own columns in the report builder.
Self-serve, per page processed. Start with the free trial: 200 pages for 90 days, no card (one credit is one page). Then keep going pay-as-you-go with credit packs, which stay valid for 12 months, or move to a monthly plan (Standard, Pro or Firm) once you know your volume. Enterprise is custom. See full pricing.
Free 90-day trial — 200 credits, no card.