Case studies
Outcomes we are allowed to talk about.
Concrete numbers, named clients where possible, honest retrospectives. If a project is covered by an NDA we say so instead of faking anonymised claims.
Client · Mid-sized DACH insurance group
Benchmarked 18 AI opportunities and shipped 3 production pilots in 90 days
A DACH insurance group was sitting on a 200-page 'AI strategy' from a Big-Four consultancy that hadn't produced a single production system. We ran a 1-week strategy sprint, benchmarked 18 candidate opportunities, and shipped three live pilots in the following 90 days — all inside an EU AI Act governance frame the client's board actually signed off on.
- 18
- Opportunities benchmarked
- 3
- Pilots live in production
- €420k
- Projected Y1 savings
Client · Swiss DTC premium brand
Rebuilt a sluggish storefront into a sub-second Hydrogen site — conversion up 2.4×, AOV up 38%
A Swiss premium D2C brand was losing customers on a slow, bloated Shopify theme. In three weeks we shipped a custom Hydrogen storefront with a faster checkout, an editorial feel, and Core Web Vitals deep in the green — conversion went up 140%, LCP came down 62%, and average order value climbed 38%.
- +140%
- Conversion rate
- −62%
- Largest Contentful Paint
- +38%
- Average order value
Client · German B2B industrial manufacturer
Replaced three legacy tools with one custom CRM and cut the quote cycle by 40%
A 180-person German manufacturer was running its sales operation across an Access database, a spreadsheet, and a shared Outlook inbox. In five weeks we designed and shipped an internal CRM tailored to their exact quote-to-order flow — one system, one login, 40% faster quotes, and roughly €120k of recovered time per year.
- 5 weeks
- From kickoff to UAT
- 3
- Legacy tools retired
- 40%
- Faster quote cycle
Client · DACH FinTech scale-up
Cut invoice processing time by 78% with a human-in-the-loop AI agent
A German FinTech was drowning in 4,000 supplier invoices a month. In four weeks we shipped an LLM-powered triage agent with extraction, validation, and a clean human review UI — cutting handling time from 14 minutes to 3 per document and freeing 180 operator hours every month.
- 78%
- Faster per invoice
- 180 h
- Operator hours saved / month
- 99.2%
- Extraction accuracy
Your move
Let's make your software feel inevitable.
Tell us what you need. We reply within one working day — with a real opinion, not a sales pitch.