AI automation case studies in Luxembourg

Most AI articles trade in promises. This one trades in delivered work. Below are three real projects 20 More built, with the outcomes their teams actually measured. For the wider context, see our guide to AI implementation in Luxembourg.

Results at a glance

M&A advisory firm: from 20-40 hours to ~40 minutes per document

Challenge

Producing M&A documentation was hugely time-consuming. Each document took between 20 and 40 hours to assemble — drawing together scattered figures, reconciling inconsistent data, and formatting everything to a professional standard. It was skilled, expensive work, repeated deal after deal, and it bottlenecked the firm's capacity to take on mandates.

What we built

We built a set of Claude plugins paired with a consolidated, clean dataroom in which the underlying data is normalised. Rather than starting each document from a tangle of source files, the team now works from a single, structured source of truth, with the plugins handling the heavy lifting of drafting and assembly.

Result

The firm now produces full M&A documentation in about 40 minutes, down from 20-40 hours per document. That is a time reduction of roughly 98% on the headline task — turning days of work into the better part of an hour, and freeing senior people to spend their time on judgement rather than assembly.

Clubhouse: automating the supply-side bottleneck

Challenge

Clubhouse is a student-housing agency that places students in apartments across Paris, Berlin, Madrid and London. Student demand was never the problem — it was always there. The constraint was supply. Sourcing apartment listings was done manually, took hours every single day, and was the team's most tedious bottleneck: the work that stood between them and a bigger portfolio.

What we built

We built an automated sourcing tool that continuously monitors newly posted rental listings across the major platforms in each of the four countries. It consolidates everything into a single dashboard and attaches the corresponding landlord or agency contact details to each listing — in effect, a streamlined lead generator purpose-built to grow Clubhouse's apartment portfolio.

Result

The sourcing process is now fully automated. Hours of daily manual work have been eliminated, freeing the team to focus on what actually moves the business — closing landlords — and letting Clubhouse scale operations on the supply side rather than being throttled by it.

FILIA: a full Luxembourg property portal, built solo

Challenge

Luxembourg's benchmark property portal, athome.lu, is run by a well-resourced company with a dedicated engineering team and tens of thousands of listings. Building a credible, full-featured alternative would normally take a whole IT department and many months.

What we built

FILIA was built the modern way: one founder, hands-on, amplified by AI-assisted development — product, architecture and engineering in a single seat. AI-powered search, multi-agency listing aggregation and a polished public portal, shipped rather than prototyped.

Result

filia.lu went live as a real, full-featured Luxembourg property portal — built essentially solo in about three months. The headline numbers tell the story:

• 3 months from zero to a live portal
• 1 founder, not an IT department
• AI-assisted development as the multiplier that made it possible

"I built filia.lu essentially on my own, in about three months. The point isn't that I'm a one-person army — it's that AI-assisted development has changed the economics of building software. A serious portal that used to need a full IT team and a year is now something a focused founder can ship in a quarter."

— Laurent Tousch · Founder, 20 More

It is proof, in production, that work once reserved for a large IT team can now be delivered by one founder with the right AI workflow.

What these projects have in common

None of these is a generic, off-the-shelf deployment. Each is a custom system built around how the client actually works — the M&A firm's documents, Clubhouse's four-country supply problem, FILIA's listings. That is where the measurable gains come from: not from AI in the abstract, but from AI pointed precisely at the task that was costing the most time. Read more about how EU companies replace manual admin work with AI.

Could your most tedious task be the next case study?

Every project above started the same way: with one painful, repetitive task that ate too many hours. If you can name that task in your own business, we can usually scope an automation for it.

At 20 More, we design, build and run the AI systems that produce results like these — EU AI Act compliant, with up to 70% Luxembourg state funding for eligible SME projects.

Schedule a 30-minute consultation to discuss what we could automate for you.