EBTS - AI AUTOMATION
ebts logo

Last updated: 28 April 2026

How EBTS reduces invoice workload by 25% with AI

EBTS, a Belgium-based provider of 24/7 European roadside assistance for trucks and commercial vehicles, struggled to manually process thousands of multilingual invoices each year. To address this inefficiency, EBTS partnered with ACA Group to implement an AI-powered invoice processing solution using automation and machine learning. This approach reduced manual workload, improved data accuracy, and accelerated invoice handling, enabling more efficient financial operations at scale.

Read more
Share this case:
01

Challenge

Manual invoice processing across 16 languages created delays, errors, and heavy workload for EBTS’s 24/7 operations.


02

Solution

A fast, low-disruption rollout using AI as a backend add-on, combined with historical translations and LLMs for accuracy.


03

Approach

AI-powered system for automated translation, smart data extraction, and built-in validation, integrated into existing workflows.


04

Result

25% workload reduction, faster processing, improved accuracy, and ROI expected within one year.

ebts logo

About EBTS

EBTS is a Hasselt-based company specializing in 24/7 European roadside assistance for trucks and commercial vehicles, has taken a big step forward in efficiency. Facing the challenge of manual invoice processing, multilingual invoice management, and time-consuming translations, EBTS teamed up with ACA Group to implement AI-powered invoice automation.

Project overview

Together with EBTS, ACA Group built an innovative solution that simplifies invoicing from start to finish: 


  • AI-powered translation: Automatically translates invoices into English and then into the client’s language
  • Smart data extraction: AI picks out key details like quantities and amounts while ignoring irrelevant info.
  • Automated accuracy checks: Ensures correct amounts and currencies through built-in verification.
  • Smooth backend integration: Works as an add-on to EBTS’s existing setup, minimizing disruption to EBTS’s existing systems.
  • Time and cost savings: The system takes 25% of the workload off a full-time employee’s plate and is expected to pay for itself within a year.

Take a look at EBTS

Challenge

Manual invoice processing & language barriers in finance

EBTS's 24/7 call center handles a high volume of roadside assistance calls across Europe.

"We essentially provide emergency roadside services for trucks," states Charles Vanmarsenille, Sales and Network Director at EBTS. "When a truck driver experiences a breakdown, we help them find the nearest garage and coordinate the entire process, including invoice handling and communication with both the customer and the driver, in 16 different languages," explains Charles Vanmarsenille, Sales and Network Director at EBTS.

However, the manual invoice processing system was becoming a major bottleneck. "We relied on tools like Google Translate, which was time-consuming and prone to errors," Charles adds. Recognizing the need for automation, EBTS set out to transform its invoicing process.

Solution

 A custom AI system with advanced AI models

ACA Group built a custom AI system, using Microsoft Azure and advanced AI models to automate translations and analyze invoice data. "The AI extracts key data and eliminates irrelevant information, which saves us an enormous amount of time," Charles notes.

"The solution functions as an add-on in our backend, without requiring us to change our existing systems", Charles explains. This made the implementation quick and straightforward, allowing the team to continue working with the technology they were already familiar with.

ACA delivered the project in just 22 days, even though 28 to 39 workdays were initially estimated. "That was a pleasant surprise", says Charles. "The solution was delivered faster than expected and within our budget."

Alexander quote

Smart use of Large Language Models (LLM’s)

The invoices requiring translation are often technical and complex, containing industry-specific jargon from nearly all European languages, which can make their interpretation difficult. To ensure high-quality translations, ACA used a smart setup with LLM’s. 

“We primarily rely on historical translations that have already been done”, says Alexander Frimout, Lead AI Innovation at ACA Group. “If a technical term was previously translated in a specific way by EBTS experts, we preserve that translation. When no historical translation exists, we utilize Large Language Models like ChatGPT to provide an accurate translation.”

“These language models are highly effective translation tools, far more powerful than solutions like Google Translate”, Alexander explains. “By combining them with EBTS’s expertise, we have created a system that is both highly accurate and consistent.”

In just two months, EBTS achieved a 25% reduction in workload for their invoice processing team. "Our employees are now free from repetitive tasks and can concentrate on higher-value activities," Charles says. "We estimate that by the end of 2024, we will have saved the equivalent of one full-time employee." Furthermore, the solution is on track to deliver a full return on investment within one year.

"In just two months, EBTS achieved a 25% reduction in workload for their invoice processing team. "Our employees are now free from repetitive tasks and can concentrate on higher-value activities," Charles says. "We estimate that by the end of 2024, we will have saved the equivalent of one full-time employee." Furthermore, the solution is on track to deliver a full return on investment within one year."
Charles Vanmarsenille
Sales and Network Director
Charles VanmarsenilleSales and Network Director
ebts logo
Charles

Future

Building on the success of this project, EBTS and ACA Group are now exploring further AI enhancements, including improved UI/UX and additional automated checks. "We have ideas for implementing additional checks, such as identifying unauthorized expenses," Charles reveals. "The potential for further AI integration is significant."

"We highly value the open communication and excellent support from ACA Group," Charles concludes. "The solution delivers substantial time and cost savings, and the ROI is very fast. We look forward to continued innovation with ACA."

Let's work together!

Interested in finding out more about AI for your business?

Contact our AI Expert Alexander Frimout

Alexander Frimout
Let's work together!

Interested in finding out more about AI for your business?

Contact our AI Expert Alexander Frimout

Alexander Frimout

Related cases

Discover how other companies transformed their digital solutions! Check out our related cases and see how we create innovative, user-friendly, and future-proof applications:

CM, a Belgian health insurer, migrated its Atlassian environment from Data Center to Cloud, with integration into Azure Entra ID as a requirement. The technical integration through Atlassian Guard was easy to set up. The real challenge was operational: deciding in which platform licences, groups and permissions are managed. ACA Group listed the decisions to make, guided and advised CM, and made sure all stakeholders were informed and aligned. The result: licences managed in Azure Entra ID, permissions in Atlassian groups, and an enterprise-grade identity and security solution with clearly defined roles and responsibilities.

CM
CM Christelijke Mutualiteit logo

Jessa Ziekenhuis, a hospital in the healthcare sector, was struggling with C2M: outdated Electronic Health Records software with rising maintenance costs, growing unavailability and a stalled feature backlog. Together with ACA Group, the hospital built GEMS (Global Electronic Medical System) on the ACA Platform, with a joint development team and a strong focus on UX. After a phased rollout, GEMS fully replaced C2M by March 2018. It now serves about 2,500 users every day, with fast response times and no unplanned downtime.

Two Jessa Ziekenhuis care providers wearing face shields and protective gowns
Jessa Ziekenhuis logo

When Stellantis  inherited an outdated Jira Server environment, simply moving everything to the cloud was not the answer. Years of duplicate fields, inconsistent workflows and legacy configuration would have moved with it. They wanted something better: a clean Jira Cloud environment without losing years of valuable history or disrupting its teams.

Together with ACA Group and Exalate, Stellantis migrated around 150.000 Jira issues in just 25 days, while cleaning up and restructuring its data along the way. 

blue orange background
Stellantis logo
Until mid-2024, a major Flemish government organization ran its operations from an extensive on-premise environment that was reaching its limits: outdated hardware, tightening security requirements, and little visibility into IT costs. ACA Group delivered a large-scale migration project, bringing more than 80 applications to Microsoft Azure-cloud with minimal downtime. The result: a safer, scalable, and fully transparent platform ready for tomorrow's innovations.
Developed by Microsoft Azure
VIAS institute wanted to reduce its reliance on government-funded projects by developing scalable digital products for new B2B and B2C markets. ACA Group conducted market research, stakeholder workshops and business case validation to create a strategic digital roadmap with five value streams. The roadmap provides VIAS with clear investment priorities and a solid foundation for sustainable digital product development.
vias header case
vias logo
The Flemish government (public sector) faced fragmented real estate data spread across multiple agencies, making it difficult for citizens and stakeholders to access certificates, permits, and reports. ACA Group developed the Gebouwenpas, a scalable digital platform built on Liferay DXP with Spring Boot and Angular, centralising all building data into a single digital passport. Using an agile Scrum and Kanban approach, the solution delivers clear insights, secure access, and a future-proof system for both residential and non-residential properties.
OVAM_dossierdoolhof
OVAM-Leeuw-engels_groen-letteromtrek

Share this case with your team members!