Lightning fast, AI-powered document processing for your logistics business 

Clients can expect up to 90% cost reduction in cases with high document processing demand

Lower processing costs

Supporting your human operators with AI-powered document processing algorithms increases capacity.

Reduced discrepancies

Greatly minimised chance of mistakes in high-volume document processing scenarios.

Fast feedback loops

Eliminate situations where collaborators send unintelligible documents. Accept or ask to retry on the spot.

Faster invoicing

With faster document processing and gathering invoices can be sent or paid much faster, improving cash flow.

How do we achieve these?


All documents received are split into categories such as: CMR, Weight ticket etc. It learns as employees process documents that are unfamiliar to the AI and other past identifications. 

Data Matching

All documents received are run against a matching algorithm with data from your system. The result is Found/Correct with varying confidence percentages, enabling further manual or automated actions. 

High accuracy and self-learning algorithm

Overall, it recommends manual checks to only 10% of the documents we’ve processed until now - 90% were correctly identified. This represents a huge reduction in time spent for people doing the checks. 

Easy to integrate into your systems

No matter the systems you use, the same performance is guaranteed. As a one-stop-shop service provider, we can even go further and improve the systems around it for even more benefits to your business.

Use case examples

In the first phase, collaborators upload different documents to your processing portal. DOX DetectTM checks all uploaded files and classifies them by type (e.g. CMRs and Weight Tickets).
According to specific rules for each type, it accepts or rejects documents almost instantly. Thus, the person who uploads a document knows if it is accepted or not, on the spot. If not, they can try again with a better picture/scan until they are accepted.

Portal upload for classification

The second phase consists of information matching. This means that DOX DetectTM checks the information within unprocessed documents against your systems, then signals if specific items are found in your database (e.g. shipment number, load address, unload address, order number, load and unload reference etc.). Each such item can be given a degree of importance and at the end, it outputs an average score that signals a match.

Section for data matching

Why choose DOX Detect?

Tangible benefits


1 Facliei Street
410181 Oradea, Romania


Since 2008, Qubiz builds new products and software solutions that help small, medium and large companies grow.

© 2020 Qubiz.

All rights reserved.


6 Mihai Eminescu Street
400033 Cluj-Napoca, Romania


Read the

Read the

Privacy Policy
Cookie Policy

A product of

Schedule a Demo

Fill out the form below and we'll arrange for a demo. If you would like to use your own documents, please prepare at least 10 files (e.g. CMRs)

Botond Gagyi

Strategic Partnerships Manager

Phone: +49(0)170 48 53 770


Take a sneak peek. See it in action.

"With DoxDetect we are improving the quality of scanned documents. Before using DoxDetect, our suppliers could upload any document, as long as a file was uploaded. Without assessing quality criteria, the documents were automatically stored on a local disk. Customers sometimes requested a copy of the documents later in the process. At that moment, we found out we had no representative documents available. Today, the quality of a document is assessed directly after uploading. We avoid incomplete or unrepresentative documents upfront. This saves time and hassle for all parties."

Pirke van den Elsen

Chief Digital Officer