what we do
We tailor cutting edge Natural Language Processing (NLP) to your business
Digitizing Financial Statements
The Balance Reader is a platform, that uses Artificial Intelligence to digitize financial statements. It takes annual reports or financial statements as input and outputs the financial data in a standard data format.
Combining OCR, image recognition models and text classification we are able to identify relevant pages, classify different types of financial statement and extract the financial data.
Extracting data manually from annual reports is a slow and cumbersome way to get financial data. With the balance reader, this task is automated.
Redactify is a software that generates an anonymized version of a document.
With the help of artificial intelligence, our algorithms identify words as people, places, businesses, or similar entities. This enables us to automatically recognise personal information in a document and redact it. The redaction is secure, so that the anonymised words cannot be restored.
Finding personal information in an unstructured text is a very long and tedious process if done manually. With Redactify this process is done in a matter of seconds.
Full Document Processing
As part of the project, we extracted relevant content from contracts and transferred it to a database.
By combining various AI methods, domain expertise and Natural Language Processing, we were able to extract content from the contracts and store it in a structured form.
The advantage for our customers is that the contents of the contracts can now be accessed more easily and quickly. Due to our standardization our customer can compare their contracts and gain insights across their portfolio.
Form field redaction
Anonymizer AU consistently redacts sections of documents.
Using state-of-the-art object detection we identified text boxes on forms. We then anonymize content of the boxes by blurring the image in that section while keeping the rest of the document intact.
One of the biggest assets of big corporations has become their data. But in order to harness this potential without privacy issues companies need to reliably remove personal & sensitive information from them.
In this project we build an intelligent input management which classified inbox documents and organized them into target folders.
By combining object detection and text classification we were able to build a system that categorizes the documents correctly for further processing.
The use of our software relieved employees of the problem of tedious and time consuming manual sorting of their inbox.
Customers & Partners
Have a project in mind?
We can help you to identify suitable projects, test their technical feasibility and successfully implement them.