rovatix · AI-based document processing

Stop retyping your documents.

rovatix extracts data from PDFs, scanned documents and photos. Check the result, correct it where needed, then use it in the format you need.

The export, import or integration your own software needs is agreed separately.

  1. Document
  2. Check
  3. Expected output
Screenshot of rovatix: a handwritten invoice and the fields read from it.
The original document and the extracted data side by side.
Screenshot: a printed purchase order and the summary rovatix made from it.
A printed purchase order and the summary made from it.
Screenshot: a delivery note's line items in a table.
The delivery note's line items in a table.
Screenshot: a Hungarian job sheet's data in English.
A Hungarian job sheet, translated into English.

We start where data is typed in by hand today.

Delivery notes

Header data and line items read, checked and prepared for further processing.

Orders

The data of incoming documents put in order, so less has to be copied by hand.

Job sheets

The recorded data reviewed and processed, with a human check matched to the quality of the document.

We test how well it works, and how much checking it needs, on your own documents.

From the document to checked data.

Hand over the document

We start from a PDF, a scanned document or a photo.

Data extraction

The system recognises the extracted data and shows it in an orderly form.

Check and correct

You review the result and correct any wrong or missing values.

Expected output

We prepare the data in the agreed format. Importing it into your target system, or a direct integration, is agreed separately.

A free two-week pilot, with the scope set in advance.

On one chosen document type and workflow, we see how rovatix can serve your task. Before we start, we agree in writing on the sample documents, the volume, the data you need, the output and what counts as success.

The free pilot covers testing the agreed documents and showing the agreed output. A full roll-out, custom system integration and further development are part of a separate agreement.
  1. Before we start

    We agree on the task and the samples needed. The two weeks start on the agreed date, once the samples and information we need are available.

  2. Week one

    On the sample documents we show the extraction results and the cases that need checking.

  3. Week two

    Together we review the results and the agreed output, and talk through the terms of further use.

  4. Deciding on the next step

    For continuing, we make a separate offer. The pilot does not turn into a paid service by itself and does not commit you to an order.

Frequently asked questions about rovatix.

Which documents can we start with?

For example delivery notes, orders and job sheets, as PDFs, scans or photos. We agree on the exact documents before we start.

Does it work with handwritten documents?

Handwriting and image quality affect the extraction. We test the result, and the human check it needs, on your own samples.

Is a human check needed?

It is worth checking the result and correcting it where needed. During the pilot we also see which data needs particular attention.

Can you load the data into the software we already use?

That depends on the software's import and integration options. We check those first, then agree on the right output or the integration work.

What happens after the pilot?

We talk through what we learnt and make a separate offer for further use. It continues only under a separate agreement.

How do we hand over the documents?

Before a demo or a pilot we agree on how documents are handed over, on access and on the data processing terms. You do not need to send any documents when you first get in touch.

Show us which document you copy by hand today.

For the first conversation, the document type, a rough volume and the name of the software where you want to use the data are enough.

An unhandled error has occurred. Reload 🗙