In the world of digitalization, image recognition plays an important part. A lot of the forms people want to digitize contain images, text or handwritten notes that bear essential information. During the past years, lots of technologies were developed to support this transition. When it comes to the scanning of these forms, the emphasis was put on OCR as a tool to convert images to text.

OCR is great when it comes to capital letters, or printed text. But it does not work well when the documents contain handwritten information, or for example when a matrix printer is being used. This specific issue can be resolved using Singer.


Singer is a component that companies can use to transform handwritten or matrix printed text and codes into text objects (e.g. JSON) that can be sent to third party receivers. This component is able to do such a transformation because of its AI capabilities. Based on training data, and feedback of operational scanning, Singer is able to learn and identify line patterns as text.

Singer focuses on short pieces of information (like codes, dates, numbers,…). It is not intended to read entire paragraphs of text. It is therefore extremely well suited as an element in the digitalization chain. At first, the component will cut out the image blocks that need to be scanned. As a next step, these blocks are sent to the detection algorithm, which in turn will respond with the interpreted text.

Since Singer is an AI solution, it needs to be retrained. New patterns will pass through the system, and by means of a validation/feedback system, these new patterns and labels will be ingested into the Singer learning algorithm. When used via our SaaS offering, this quality monitoring will be provided out of the box.


Singer can be offered via 2 means :

  • Software as a Service (SaaS) : 
    • The entire Singer solution will be hosted and operated by Vectr. The customer will be able to connect using API endpoints
    • Monitoring will be done by the Vectr team
  • On-prem 
    • Singer will be run on the infrastructure provided by the customer
    • the customer will bear responsibility for keeping the system trained

Implementation possibilities

As mentioned above, Singer will be offered to the customer by means of API endpoints, either on prem or via the SaaS offering. The customer is able to choose if he wants to perform the orchestration of the activities themselves, or they can opt to use the Singer out-of-the-box orchestration.

For each implementation it is advised that the customer provides a (labeled) training data set. This way, the solution will be performing at a higher level at the beginning of the project. If this cannot be done, it will start learning from scratch. In the end, Singer will be equally performant but it will take more time since it has to learn everything from production activities.


Singer is well suited to be used in the digitalization of manual document flows. It can be run individually or as part of a set of digitalization tools. Individual uses can be the scanning and interpretation of :

  • logistical documents (CMR, order sheets, …)
  • quality documents (automotive, health industry, …)
  • banking documents 
  • public service docs (travel documents, border control, …)

An example where it is part of a chain of tools can be where the first tool scans the document, then sends it to Singer for analysis. Singer sends the output to a validation tool. Depending on a positive or negative validation, the information a trigger with this information is sent to a CRM or ERP system.

Reference projects

One of the best reference projects to use, is the one that Vectr has done together with the Christelijke Mutualiteiten. The full project case , with supporting movie clip, can be found here :


Is Singer out of the box available ?

No, there is always an amount of time that Vectr needs to setup the detection algorithm, depending on the number of fields you want to scan.

Does Singer come with pre-trained models ?

Only limited trained models can be offered (date field for example). Since the training is done specifically on the input types, it is not possible to provide other pre-trained models

What is the licensing model ?

Depending on the number of fields you want to scan, the licensing tier is selected. <<insert more info about licensing>>