From Idea to CE Mark: Building a Class IIb SaMD for Type 2 Diabetes
How I turned clinical ambition into a compliant, market-ready digital therapeutic for Type 2 diabetes in under 12 months.
Context
One of the top 10 global pharmaceutical companies approached me with a clear but ambitious goal: to build a mobile-first digital therapeutic that could personalize lifestyle interventions, track biomarkers, and adapt to real-time data from patients with Type 2 diabetes.
They had world-class expertise in diabetes care but limited internal capability to develop regulated digitalducts. I was brought in to lead the product and engineering effort—from defining the MVP to achieving CE certification. We had 12 months, a hybrid team, and zero margin for error.
Problem
Type 2 diabetes requires continuous behavioral adaptation and day-to-day self-management. Most digital tools on the market fall short. They log data, but they don't support therapeutic decision-making or drive behavioral change in a meaningful, clinically validated way.
Our objective was to build a Class IIb Software as a Medical Device (SaMD) that could do exactly that. Which meant dealing with clinical validation, regulatory scrutiny, and rigorous technical quality—without compromising speed or user experience.
Solution
I built our strategy around three principles: regulatory-driven architecture, evidence-backed design, and lean, compliant execution. Every component was designed with IEC 62304, ISO 14971, and MDR in mind—from data models to UI flows. We collaborated closely with endocrinologists and behavioral scientists to embed clinical logic directly into the product. And we ran agile sprints layered with design controls—clear documentation, traceability, and audit readiness from day one.
What we delivered was a mobile app that integrates with CGM devices, helping patients visualize glucose trends and receive contextualized behavioral nudges. We built a clinical engine that blends rule-based logic and machine learning to deliver personalized insights and detect risk patterns. To support all this, we developed a compliance stack that embedded traceability into Git workflows and auto-generated most regulatory documents as part of our development pipeline.
Go-to-market
To accelerate the CE marking process, I worked closely with regulatory teams and QA stakeholders. We conducted light-touch internal audits every sprint, which made the final compliance review almost uneventful. Our git pipelines automatically linked every code commit to a requirement, risk, or test artifact, providing end-to-end traceability with minimal manual overhead.
Once certified, we launched pilot programs in select European markets to begin collecting real-world evidence.
Outcome
In under a year, we went from idea to a CE-marked Class IIb SaMD that’s now actively helping patients with Type 2 diabetes achieve better outcomes.
The real lesson? Proving that regulated medical software can be built at startup speed—if you architect for compliance from day one.