
LIST Gruppe: AI with a real starting point
The LIST Gruppe didn't want to run AI as a tool debate, but to show what's possible on a real work process. Together we set up an AI partnership that starts with one prioritized use case and turns it into a solid basis for deciding on the wider rollout.
1 min read
Industry
Construction & Building (general contractor)
Format
AI partnership with an evaluation phase
Starting use case
Incoming invoices · VAT review
Way of working
Prioritization in the jour fixe · 1- to 2-week sprints
Problem
Many possibilities, but no proof in daily work yet
In large construction companies, AI ideas pile up quickly: review documents, find knowledge, support estimating, ease administration. Without clear prioritization, that easily turns into a long list. LIST didn't need another overview of possibilities, but a concrete starting point that's domain-relevant and can be measured in operations.
Solution
Start with a workflow that counts
The first focus is on incoming invoices and VAT review — a process with volume, rules and tangible potential for relief. Instead of designing for months, we work in short sprints: sharpen the use case, build the solution, validate with domain expertise, document the results, and hand them over to the organization.
- Prioritization — Not all ideas at once, but the next domain-sensible decision in the jour fixe.
- Delivery — AI Engineer, Adoption Manager and partner work close to the business unit, so the solution and the reality of work fit together.
- Handover — Documentation, operating notes and knowledge transfer are part of the project, so the use case doesn't end as a prototype.
Impact
AI becomes a basis for decisions
The approach takes AI out of the abstract. LIST sees on a real process what quality is achievable, which organizational conditions are needed, and where a larger partnership pays off. That's exactly how AI adoption should start in the mid-market: not with a slide full of possibilities, but with a work process that runs better after the project.


