Computer with artificial intelligence process for processing PDF data with AI

Each week, TYT Group receives approximately 500 container transportation order requests by email in various formats, including PDFs, scanned documents, screenshots, and content within the email body. Although each document contains the essential information needed to create an order, the data is structured and worded differently each time, depending on the forwarder or client.

The main challenge? Only an experienced person could process these requests by:

  • Opening each email one by one and sorting the requests between import and export.
  • Decoding each document to identify and gather the required information.
  • Quickly identify missing information to retrieve it from the forwarder, client or another source.
  • Creating the orders and manually transcribing the data into the transport management system (TMS).

This manual process wasn’t just time-consuming; it monopolized a key resource who was solely dedicated to processing orders, creating a bottleneck for TYT Group’s growth. Furthermore, a single data entry error could lead to delivery delays and significant financial penalties if a container were held up at the port.

The solution: an AI agent that understands all document types

The project’s goal wasn’t to replace human expertise, but to amplify it with an AI virtual agent. We developed a proof of concept for a custom AI agent designed to seamlessly integrate into TYT Group’s existing processes:

  • Read all document formats using optical character recognition (OCR).
  • Understand the structure of documents (like tables) and the information within them, whether from a PDF, text within an email body, a scanned document, or other formats.
  • Contextualize each piece of information to interpret and format it uniformly.
  • Extract the required data and standardize its presentation according to TYT Group’s business rules and TMS requirements.
  • Convert units of measurement to standardize them and prepare the extracted data coherently so that an employee can validate the information in the TMS before the final order entry.

Maximizing human efficiency with AI

Thanks to this proof‑of‑concept AI agent, manual and repetitive data entry was transformed into a much faster and more efficient supervision and validation step. This has allowed us to demonstrate that it is possible to:

  • Reduce order processing time by automating repetitive manual tasks, freeing up employees to manage exceptions and focus on strategic activities.
  • Limit the risk of manual entry errors that lead to delivery delays and penalties.
  • Unlock the company’s growth potential by enabling a higher volume of orders to be processed without increasing operational costs.

Helping your business grow with custom AI agents

This collaboration with TYT Group proves once again that AI is a powerful performance lever. By combining artificial intelligence with our custom development expertise, we create tangible solutions that augment people’s ability to create value rather than replacing them.

Are your manual processes holding back your growth? Contact our team to see how we can integrate artificial intelligence into your business processes, enabling your company to grow and empowering your employees to focus on strategic, high‑value tasks.

Got a project in mind?

Want to start a DevOps project? We have all the technical expertise you need under one roof to support your initiatives!

Write us Our projects

An Nmedian talks to another in his office at Nmedia
Discover job offers