Problem
Statement
Everyone wants AI — hardly anyone has a plan
AI tools are provided but barely used internally because the use cases are missing. On top of that, there are no clear rules for their use to prevent shadow AI. And the EU AI Act training obligation already applies.
An AI rollout framework: use case, rules, enablement, rollout — step by step.
The project is moving — just not forward
Milestones keep slipping, the steering committee meets without deciding, the budget shrinks faster than progress is made. Nobody wants to stop the project, but nobody knows how to get it moving again. The longer this lasts, the more expensive every restart becomes.
Turnaround in three steps: an honest stock-take, a revised scope, clear accountability through to go-live.
Pilots don’t scale
What one committed person builds stays with them. Without roles, an approval path and clear accountability, an individual initiative never becomes a standard for the whole administration.
Governance that enables rather than prevents: roles, approval, reuse.
Every rollout is a new project
Software is introduced as a one-off project: tender, rollout, project end. For updates, new modules and the next specialist application, everything starts from scratch — FitConnect alone quickly means around 100 web forms, each on its own. The result is project fatigue.
A repeatable rollout framework: set it up once, use it continuously.
Vibe coding is happening — allowed or not
Staff are already building interfaces and small applications themselves with AI, often bypassing IT. A ban does not stop it; it only makes it invisible.
Make in-house builds visible, review them and move them cleanly into operations.
Knowledge lives in heads and folders
The retirement wave meets scattered storage: templates in SharePoint, emails and local folders, and nobody knows for sure which version applies. Whoever leaves takes their knowledge with them.
A local AI knowledge platform built on existing language models: answers with sources, no cloud.
The EUDI Wallet is coming — but where exactly?
From January 2027, public bodies must accept the EUDI Wallet. Hardly any town hall knows which of its own services are actually affected. In one example town in North Rhine-Westphalia: 268 services under its own responsibility, only 31 already wallet-ready, 162 requiring action.
An assessment cascade instead of panic: a prioritised work list up to 2027.
Patient intake still on paper
Appointments by phone only, medical history forms on a clipboard, data typed in by hand, exercise plans sent home on paper. It ties up the front desk while Germany is short of more than 12,000 physiotherapists. Practices that build their own online forms quickly end up with US tools — a risk for health data under Art. 9 GDPR.
The patient journey made digital, step by step: appointment booking, online medical history, direct transfer into the practice management software and follow-up care — GDPR-compliant.
The telematics mandate is coming — and nobody has time for it
Connector, card terminal, institution card (SMC-B), electronic health professional card, KIM: to connect to the telematics infrastructure, therapy practices have to order components, file applications and set up the practice software — on top of a fully booked treatment schedule. The statutory TI reimbursement covers the costs — but only once everything is in place.
A TI roadmap for the practice: what to order, in which order, and how the practice then uses electronic prescriptions and the ePA.
Treat patients, not paperwork
Billing, bookkeeping, social media, correspondence: in small practices the owner often handles all of it after hours. AI agents can take over these routines — once it is clear which data they may see and what the German Healthcare Advertising Act (HWG) allows in marketing.
AI agents for everyday practice work: one use case at a time, with guardrails for data protection and healthcare advertising rules.
When IT projects stall
Rollouts that have not moved for months: the software is bought, the wards keep working the old way, and KHZG-funded projects face penalty deductions. Between department, IT and vendor it is unclear who decides what — so nobody decides.
Turnaround with an honest stock-take: what holds, what gets stopped and who is accountable through to go-live.
AI pilots stay pilots
A chatbot here, speech recognition for documentation there: hospitals and practices run AI pilots, but rarely beyond the first ward. Rules for health data, approval paths and clear accountability are missing. And the training obligation under the EU AI Act already applies.
An AI scaling framework: use case, rules, enablement, rollout — from pilot to regular operation.
HIS plus 700 subsystems
The real madness inside a hospital: system landscapes grown over years, interfaces only one person understands, and operations that spend more time firefighting than improving. Every outage takes time away from patients.
Turnaround in IT operations: put the system landscape in order, assess and consolidate subsystems, set up stable operating processes.
Cases
Innovation Hub
Service
Journey
Concept
Understand, assess, decide. Self-checks, compact analyses and structured pathways create clarity within days on whether — and where — digitalisation pays off.
See Innovation Hub ↑Build
From idea to working approach. Concrete implementation, pilots and prototypes — with clear ownership and a deadline.
See Cases ↑Run
Sustainable in daily operations. Support in operations or ongoing collaboration, so digitalisation doesn't stall again after kick-off.
Research & Teaching
Sharing what works. Talks, teaching and applied research — so good solutions don't stay with just one client.
How to reach me
First contact
Assess it yourself, no obligation, or reach out directly with an idea — via a self-check in the Innovation Hub or a short message.
Orientation call
20 minutes, no obligation. We work out together whether it fits.
Proposal & start
A tailored proposal and a start together.