Motulus

The client

Motulus is a Belgian scale-up that builds smart optimization software for the aviation industry. Drawing on their expertise in mathematical algorithms, software engineering, and high-performance computing, they develop proprietary solutions for areas including crew scheduling, pairing & roster production, tail assignment, and network optimization. In doing so, they help airlines reduce costs, deploy their fleets more efficiently, and make their operations more robust.

With a small, highly specialized development team, Motulus builds software for complex, large-scale planning problems. This requires not only deep technical expertise but also an efficient way of working. For Motulus, AI-assisted development was a logical next step to further empower their developers and get more out of their development process.

The case

Motulus asked Euricom to structurally integrate Claude Code into the way their development team builds software. Not as a one-off training day, but as a true AI-Empowered Software Development Life Cycle: AI support that is woven into daily engineering work, rather than just sitting alongside it.

That takes two things at once. The team had to master the fundamentals, from prompting to code review and the question of where you do and do not rely on AI. And the tooling had to land inside the existing development workflow, not in a sandbox next to it.

Our approach

The track started with an interactive AI workshop, delivered by Euricom consultant and AI Coach Peter. The training was tailored to the concrete needs of the Motulus team and started from recognisable use cases out of their daily development work.

After that first training, Yorne joined the Motulus development team as a Euricom consultant. From that daily collaboration, he takes AI further into practice and helps the team discover and try out new possibilities.

That combination of training and daily guidance is what keeps AI from staying a one-off session. New questions and use cases can be picked up and tested straight away in real development work. That way, new habits keep growing after the first training as well.

That is Euricom's starting point in AI adoption: not simply introducing a tool, but building knowledge and experience that stays within the team. Throughout, the work stayed in the real codebase: the optimization software Motulus ships to airlines, not a sandbox or a side project.

The result

The collaboration between Motulus and Euricom has been running since 2024. In January 2026 it took on a new focus with the start of the AI track. Claude Code is now part of daily development work on the optimization software Motulus ships to airlines — and the effect is visible in Motulus' own Git history.

Median lead time from first commit to merge fell by 69% on the backend team, from roughly seven hours to just over two. That team did not change in size across the period, which makes it the cleanest comparison available: the same people, the same codebase, a different way of working. Work that used to linger no longer does — pull requests left open longer than a week dropped from 23% to 6%.

The frontend team now delivers 9.4 tickets per sprint against 5.5 before. A third developer joined in January, so not all of that is tooling: measured per developer, output rose 14%.

For a scale-up like Motulus, that last figure is the one with a price tag attached. Optimization engineers with aviation domain knowledge are scarce and slow to recruit, and the planning problems Motulus solves do not allow for junior shortcuts. A 14% gain per developer is close to an extra engineer's worth of capacity across the current team, without adding headcount — capacity that goes straight into the product roadmap instead of into a hiring pipeline.

Adoption has widened accordingly. What began with a first group of four developers now covers six Motulus engineers using Claude Code in daily production work, and twelve developers across the wider product team once the offshore and contract developers working in the same codebase are included.

Underneath those numbers sits the way of working that produced them:

• Six Motulus engineers using Claude Code daily in their production environment, fully integrated into the existing development workflow.
• The development team trained in a shared approach to working with AI, from prompting to code reviews.
• A regular coaching rhythm that keeps the team experimenting, learning and refining.
• Euricommer Yorne, who works alongside the team every day and helps turn AI into everyday practice.

"We build optimization software for airlines, where the hard part is never typing the code. Having Yorne on the team turned Claude Code from an experiment to our daily routine, so our engineers spend more time solving actual problems."
- Przemyslaw Klosiewicz, CTO Motulus

What began as an AI workshop has become a broader way of working, bringing together training, daily practice and ongoing guidance. The partnership will be further expanded in the coming period.

Figures come from Motulus' own Git history across 704 pull requests, comparing the six months before the AI track with the eight months after. The backend comparison is drawn from a team whose size and composition stayed the same throughout, so the change reflects the new way of working rather than added capacity.