MINT.extract — Truly refreshing document digitalization.

MINT.extract is an information retrieval engine that...

  • delivers incredibly advanced document analysis capabilities
  • goes beyond OCR (Optical Character Recognition) as it recognizes different structures and layouts of documents based on machine learning and natural language processing
  • understands semantics and uses contextual information to retrieve unstructured data in a fully automated and highly performant way

Get started by requesting access to MINT.extract

Repeat this process for any applications and use cases and drive digitalization at your pace

Extend the system with more data sources and/or add classification and validation services

Run the Machine Learning System productively in your daily processes


Label your documents with a few clicks and set the foundation for your own Machine Learning System in one single day

Train the algorithms on your documents and benefit from shared learnings to obtain the best results

Evaluate the results in real-time and control the output quality through the feedback loop

MINT.extract is in productive use in the following industries

Legal & Tax

  • Contract analyses
  • Legislative texts
  • Tax statements

Financial Industry

  • Bank statements
  • Pension certifcates
  • Product termsheets


  • Medical records
  • Drug prescriptions
  • Laboratory reports


  • Sales orders
  • Shipping certificates
  • Freight bills


  • Insurance policies
  • Quotes comparison
  • Claims handling


  • Archived files
  • Official certificates
  • Book series

Purchase Orders

A global logistics company uses MINT.extract to automatically process thousands of purchase orders. After extracting data points such as 'article number', 'amount of items', 'order date' as well as the address of the buyer, the structured data is fed via API into their enterprise system and the shipment is triggered from there.

The possibility to process orders 24/7 from different locations has boosted efficiency by a factor of 10.

Financial Documents

A wealth management firm collects and analyzes bank statements, pension certificates and insurance policies for all their clients. MINT.extract structures various document types. Peaks of workload and bottlenecks are removed and client advisors spend now more time on individual recommendations.

Customer satisfaction has risen significantly through the company's ability to generate reports swiftly.

Mail Box Classification

A Swiss retail bank receives thousands of customer documents in a central mailbox, where MINT.extract classifies the documents, identifies relevant customer information and routes them to the responsible team. As a result, tedious paper work is processed fully automated.

Response time to customer requests and costs related to manual handling have been cut massively.

Get Quarterly News about turicode

Sing up to our newsletter and receive quarterly news about interesting projects, exciting research results and more news straight to your inbox.