AI made building cheap.
Reality did not.
The Reality Sprint runs your AI product through the four gates that decide whether it survives, before you scale. When it passes, we build the version that earned it. In regulated health, the fourth gate is where most products quietly die.
The AI Product Reality Canvas, the same framework in the published post.
01
Desirability
Do people want it?
02
Capability
Can it do the job?
03
Viability
Can the business survive?
04
Practicability
Can the product survive reality?
MDR · EU AI Act · DiGA · GKV/PKV · Evidence
The problem
The expensive mistake moved.
AI made building cheap. A working V1 costs a fraction of what it did three years ago. What did not get cheaper is finding out the product cannot survive adoption, economics, or regulation. The bottleneck moved from building to building the wrong thing. In regulated health the fourth gate, legal, reimbursable, evidence of care, is where most products quietly die. It is the gate generalists skip.
The method
Four gates. The fourth decides it.
- 01
Desirability
Do people want it?
Real users with a real problem, a real workflow it can enter, and a real reason to adopt. Not assumed demand. Not assumed workflow fit.
- 02
Capability
Can it do the job?
Performance at the task, on real data, in the conditions that matter. The gap between demo and deployment is where most AI products fail.
- 03
Viability
Can the business survive?
Unit economics, pricing architecture, sustainable customer acquisition, and a clear path to margin. Not a funding dependent business model.
- 04
Practicability
Can the product survive reality?
Regulatory classification under MDR and the EU AI Act. Reimbursement pathway across GKV, PKV, or DiGA. Evidence design to demonstrate clinical benefit. Data protection architecture. This is the gate generalists skip. It is where most regulated-market AI products die.
The offer
Four sprints. Each fixed scope, fixed fee.
Most clients start with the Reality Sprint. The others attach to it.
- 01
Map Sprint
Optional start
Score a field of ideas and select the one worth pursuing. Right for teams with multiple candidate products and limited capacity to explore them all.
- 02
Reality Sprint
The core
A four gate evaluation of one AI concept. Ends in a decision ready memo and a prioritised validation plan. Two weeks. The anchor engagement.
- 03
Validation Sprint
Scoped to one target
Structured validation against users, payers, or clinicians, depending on which gate needs the deepest investigation.
- 04
Build Sprint
AI accelerated V1
An AI accelerated working V1, explicitly pre scale. Weeks, not quarters. Builds the version that passed the reality gates, not the version someone assumed would.
What you get
A decision. Not a deck.
The output is something actionable. A verdict the team can act on and the board can stand behind.
- →Go or no go decision with the reasoning laid out
- →Costed unit economics. What it takes to be viable
- →Live regulatory read across MDR, EU AI Act, and the DiGA pathway
- →Evidence read. Whether it improves care and what proof is required
- →Prioritised plan with named next steps and owners
Selected references
Ventures we built. Work we run for clients.
- 01
Echo. Pet Platform
Built solo. NFC and AI, run through all four gates. Live in market. Selected for OMR Tokenmade 2025.
- 02
Schwangerschaftsdiabetes DTx
Venture build for a foundation-funded digital health nonprofit. Daily coaching and glucose tracking, reimbursement pathway scoped from day one.
- 03
Hybrides Präventionsportfolio
Discovery across three prevention programmes for a health insurer. Rare disease, cardiovascular, chronic pain — structured as evidence-ready pilots from the start.
- 04
Alveus App
On-device pseudonymisation for clinical AI queries. Patient identifiers never leave the iPhone.
- 05
LEIA / constanze.care
AI pregnancy companion. Built and spun out. WhatsApp native, no app required.
Who it is for
Product and innovation leads in regulated health who are building or about to build an AI product, and need to know it will survive reality before they commit to scale. Someone with a product in motion and budget to reduce the risk.
Not for teams looking for a strategy deck. Not for companies that have already scaled. Not for founders at the idea stage without a product in motion.
Start here
Before you spend a build budget finding out, spend a fraction to know it deserves one.
30 minutes. No obligation. We'll tell you if the sprint is right for your situation.