We turn documents into structured data.
Invoices, orders, contracts, forms. Uncovio reads them, checks every value and delivers verified records to your systems. Accurately. Continuously. Set up in minutes.
- VendorAcme Supplies Ltd.✓
- Invoice no.INV-2026-0917✓
- Total incl. VAT€ 4,120.00✓
Works with
PDF, scan or photo. New document type? Describe it in one sentence.
Thousands of documents arrive every day. Most of their data stays stuck in PDFs, scans and e‑mails.
Uncovio reads them, checks every value and delivers the record. In minutes.
Collect. Define. Review. Deliver.
Four steps from inbox to structured record — most of them run on their own.
- 01Collect
Documents come in on their own
Connect your inbox, set a filter, done: matching attachments are picked up automatically. Or upload files in the app or send them through the API. Every page is queued in seconds.
Documents3 newDrop PDFs, scans or photos here- Invoice_September.pdfE-mail · billing@acme.comDone
- order_0917.pdfUpload · Purchase order · 2 pagesProcessing
- IMG_2291.HEICAPI · Photo · handwritten delivery noteQueued
- 02Define
Describe the fields you need
One sentence or a few example documents is all it takes. Uncovio suggests the fields and keeps every extraction consistent across suppliers and formats.
Schema · Invoice“Extract the vendor, invoice number and the total amount including VAT.”VendorAcme Supplies Ltd.Invoice no.INV-2026-0917Total amount€ 4,120.00 - 03Review
Verify every value on the page
Every field links to the exact spot in the source. Uncertain values are flagged in the evaluation UI; correct one and the self-learning engine applies it automatically from then on. No retyping.
VendorInvoice no.TotalInvoiceVerified- VendorAcme Supplies Ltd.✓
- Invoice no.INV-2026-0917✓
- Order no.PO-48213✓
- Due date2026-10-11✓
- Total amount€ 4,120.00review
- 04Deliver
Send the data where the work happens
Push verified records to your ERP, CRM or data warehouse via webhook, fetch them through the API or forward a clean summary by e-mail.
Delivered to your ERPThe verified record is already in your workflow.
TargetERP · Record 48213Latency212 msStatus200 OK · 09:24:02
- Invoice_September.pdfE-mail · billing@acme.comDone
- order_0917.pdfUpload · Purchase order · 2 pagesProcessing
- IMG_2291.HEICAPI · Photo · handwritten delivery noteQueued
- VendorAcme Supplies Ltd.✓
- Invoice no.INV-2026-0917✓
- Order no.PO-48213✓
- Due date2026-10-11✓
- Total amount€ 4,120.00review
The verified record is already in your workflow.
One API call. Any file in, JSON out.
Start a process with a document and a schema. Poll the status or let a webhook deliver the result. Typed clients come straight from the OpenAPI spec.
- POST/process/start_process— document + schema in, process out
- GET/process/status/{id}— poll until completed
- PUT/schema/{id}/delivery— webhook or e-mail per schema
# Start a process for a documentFILE_B64=$(base64 -w 0 invoice.pdf)curl -X POST https://app.uncovio.ai/gateway/process/start_process \ -H "Authorization: Bearer $BEARER_TOKEN" \ -H "Content-Type: application/json" \ -d '{"schema_id": "invoice","files": [{ "filename": "invoice.pdf", "content": "'$FILE_B64'"}]}' # → poll /process/status/{id} or wait for webhook{ "process_id": "prc_38012128618"}
Documents and extracted data stay on European infrastructure.
Data processing agreement and deletion on request.
Your documents only ever improve your own schemas.
Frequently asked questions
Answers to what we get asked most.
Can I start on my own?
Yes. Create an account, upload a few documents, define your first schema. Most teams see their first structured records within the hour. Prefer a walkthrough? Book a demo.
What file types does Uncovio handle?
Every one. PDFs, scans, photos, Word, Excel, e-mails, even handwriting — one pipeline, no per-format setup. Mix formats in a single schema and Uncovio still delivers one consistent record.
Why can’t I just let ChatGPT read the file?
A chat model gives you text. Uncovio gives you a record: your schema enforced, every field reviewed in an evaluation UI, delivered straight into your systems — on any file type, not just PDFs.
Does Uncovio use AI?
Yes. Uncovio is an AI system: AI models read your documents and fill in the fields of your schema. AI output can contain errors, so every result can be checked and corrected in the evaluation UI before it is delivered, and every delivered record carries an "ai_notice" field that marks it as AI-generated.
Does Uncovio get smarter over time?
Yes — that is the self-learning engine. Every correction you make in the evaluation UI, at any step from field mapping to review, is fed back automatically: the same fix applies to future documents without retraining or re-configuring anything. Accuracy compounds the more you use it.
Okay, so it reads the file. Isn’t that the whole job?
Reading is step one, followed by the mapping to your schema, the review in the evaluation UI, the approval, and the delivery via webhook, API, or e-mail — with the self-learning engine improving every one of those steps as you go. That is the whole job.
How do documents get into Uncovio?
Three ways: connect your e-mail inbox and set a filter, upload in the app, or send them through the API — any file type, in any language.
Can I trust you with sensitive data?
Processing in the EU, a data processing agreement, no training on your documents. Deletion on request.
I have more questions. Who do I talk to?
Us. Write to daniel.beinbauer@innomarks.de or book a 30-minute demo with the team.