AI solutions to process logistics data and integrate data information flows

ABOUT THE CUSTOMER
Digitrans is an artificial intelligence partner for companies in logistics. Digitrans products help these companies to reduce manual tasks and increase their productivity in data entry, planning, tracing assets or communicating with partners/clients.
Read the case

Digitrans specialises in building AI solutions that drive innovation in the logistics sector. Our collaboration with Digitrans showcases our expertise in AI model development and MLOps. In this case we demonstrate how our solution can improve document and data management, a critical aspect of the operations of logistics companies.

Navigating Document Complexity

The logistics industry is overwhelmed with communication in various systems, emails, and documents — generated PDFs, Excel files, scanned documents, orders in emails, etc. For Digitrans’ customers, managing this influx efficiently is a significant challenge. Traditional emails or document processing methods are labor-intensive, prone to errors, and struggle to keep pace with growing data flows. Hence, there is a need for a robust solution to assist planners in accurately extracting data and integrating it into their own logistics systems.

Our team works closely with Digitrans to develop an AI solution tailored to the logistics industry. By using advanced machine learning (ML) models and semantic AI, we created a system capable of extracting, classifying, and analyzing information from diverse document formats.

In practice, the end user forwards all documents to a data entry assistant via email. All messages received at this email address are automatically processed into a machine-readable file and sent to a preferred system, such as a transport management system or ERP.

To bring this solution to life, we use a range of open source tools:

  • PyTorch for analysing, interpreting and classifying documents.
  • Google Translate to provide accurate translations, addressing the need to process multilingual documents.
  • Tesseract OCR for converting document images into machine-readable text, enhancing text recognition accuracy.
  • Python as our development backbone, to develop the applications and the orchestration of the AI models.
  • OpenCV for image pre-processing to improve document readability and precise text extraction.
  • Nomad and Docker for the deployment of the applications and AI-Models.

Bringing AI into Production: The Power of MLOps

Developing the AI models was just the beginning. To truly improve logistics data management, we integrate MLOps practices into our solution. MLOps allows us to automate workflows, ensure continuous integration, and streamline the deployment and management of the AI models in production.


This approach improves the deployment of our AI models. It makes our solution flexible, with the ability to keep learning and improving continuously.

The Impact: Transforming Document Management

The AI system delivers:

  • Increased Efficiency: Automated document processing significantly reduces manual data entry and errors.
  • Scalability: The solution's microservices architecture, coupled with MLOps, allowed it to scale with the growing needs.
  • Continuous Improvement: MLOps practices ensured continuous deployment and monitoring, keeping the system up-to-date with the latest advancements.
  • Improved Data Accuracy: Advanced NLP and ML techniques provided precise text recognition and analysis, enhancing overall data quality and reliability.

Through this collaboration, we demonstrated how our expertise in AI model development and MLOps elevates document and data management in logistics.

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.

Do you have a question?

Nicolas Spillemaeckers
Business Developer
By clicking “Accept All Cookies”, you agree to the storing of cookies on your device to enhance site navigation, analyze site usage, and assist in our marketing efforts. View our Privacy Policy for more information.