Every project starts with a specific need. Here are some of the results.
01 / OWN PRODUCT · CUSTOMER LOYALTY
FIDEX
Bring your customers back more often.
The challenge
Stamp cards get lost or left at home, discounts are applied by hand and only reward customers who reach a minimum spend. And in generic apps, customers have to find your shop among many others.
The solution
A dedicated QR loyalty app: the customer shows a code, the shop scans it and points are applied instantly. Every euro counts, with no cards or commissions, plus data and campaigns for the business.
What it delivers
Before: a physical card that got lost, manual discounts and a minimum spend. Now: customers always have the app with them, discounts take seconds and every purchase counts, however small.
Live standings of a hill climb championship were hard to read from the course, and every official document had to be printed, signed, scanned and uploaded to the website.
The solution
A platform with live standings, category filters and driver comparison charts, plus a digital validation flow so stewards sign documents from the app itself.
What it delivers
Before: print, sign, scan and upload. Now: three clicks. And spectators follow clear, comparable live standings right from the course.
03 / EUROPEAN DISTRIBUTION · MACHINE LEARNING INSIDE SAP
DDUUEETT
Anticipating the volume of every shipment.
The challenge
DDUUEETT supplies professional healthcare, wellness and hairdressing equipment to more than 500 distributors in over 30 countries. With orders that mix items of very different sizes, they wanted to sharpen their forecast of each shipment’s transport volume even further.
The solution
A machine learning model trained on shipping history and built into SAP and the B2B shop: one button on the order and the predicted volume appears instantly. It runs on the company’s own servers.
What it delivers
In production and in daily use: a more precise forecast on every order, inside the tools the team already uses. From proposal to production in four months.
Merging race results with different formats and naming conventions to build the Cup rankings took dozens of hours of manual work each season, and some duplicates slipped through.
The solution
A desktop application that collects the files, corrects errors, standardises names, detects duplicates and generates the rankings.
What it delivers
Before: dozens of hours and duplicates that went unnoticed. Now: minutes, and duplicates are caught before publishing.
