Dr. Oliver Eidel


I'm a doctor by training but worked as a software engineer in medical software startups. Most of my time was spent at Vara Healthcare in Berlin, where we developed Machine Learning - based Radiology software.

We quickly noticed that certifying software as a Medical Device is quite tricky. And it sucks, because knowledge isn't freely available.

I left Vara to build my own business around a self-developed HR software (that didn't go well). Only then did I notice that other Healthtech startups are struggling with their Medical Device certification, too.

I got the opportunity to help other startups certify their software and resolve regulatory pain, most recently Ada Health in Berlin.

Turns out: Getting this done is pretty simple, if you know what you're doing. Unfortunately, there are a lots of consultants who make this unnecessarily complex.

I'm not very interested in regulatory consulting. I am however interested in enabling Healthtech startups access the market.

So I built this website to brain-dump my regulatory knowledge. I hope that this reduces the pain for future startups and finally enables some desperately-needed innovation in our crappy Healthcare IT Systems. Enjoy!

Companies I've worked with

Ada Health

App for personalized symptom assessment

With 9 years of research, 60 million Euro in funding and over 200 employees, Ada Health is one of Berlin's most well-known Healthtech companies. Ada's product is a mobile app which enables patients to conduct a personalized symptom assessment. It has over 10 million users and has won multiple awards.

I was brought in by Jared Cardon, Director of Compliance at Ada. My goal was to work closely with multiple teams of software engineers to optimize their development processes. Importantly, those processes should be harmonized across the company, enabling teams to ship updates regularly while ensuring regulatory compliance.

Another short-term goal was to prepare the engineering teams and review documentation for an upcoming ISO 13485 audit. They successfully passed the audit in June 2020.

Vara Healthcare

AI-based Radiology Software for Breast Cancer Screening

Vara is a Machine Learning - based software which helps radiologists assess x-ray images (mammograms) for early detection of breast cancer. It highlights healthy exams so that radiologists can spend more time looking at potentially suspicious images.

At Vara, I was one of the first employees back in 2017. I developed the first software version (frontend and backend) and recruited the (non-ML) software engineering and regulatory teams.

Together with two other colleagues, we taught ourselves how to certify ML-based software as a Medical Device. We successfully passed the ISO 13485 audit and got our software certified as class IIb Medical Device under MDD.

Exakt Health

Physical Therapy Reingineered

Physical therapy for sports injury is typically non-digital: Patients regularly visit a physiotherapist for exercise sessions. Exakt Health developed an innovative, digital-first approach by providing an app which enables patients to perform physical therapy at home.

We worked with Philip, the CEO, in setting up an ISO 13485 - compliant Quality Management System and Technical Documentation of the app which is compliant with IEC 62304, ISO 14971 and IEC 62366.


Machine Learning Biomarkers For Rare Diseases

Founded by two ex-Googlers and radiologists, imvaria is developing a Digital Biomarker Platform for Machine Learning - based, non-invasive diagnostic and therapeutic patient stratification.

While I'm certainly not an expert in FDA compliance, I did help them in implementing IEC 62304 - compliant software development processes and documentation. The main goal was to ensure that regulatory work remains as simple as possible, both ensuring compliance and good product quality, but also enabling developers to continue shipping code. I also assisted in aspects of their ISO 13485 - compliance, reducing the amount of regulatory overhead.

Their leading Machine Learning Biomarker received FDA Breakthrough Device Designation in 2020.


Computational Microscopy

When pathologists assess biopsies, they typically use analog microscopes and manually count cells. This process is slow, subjective and error-prone. MindPeak develops Artificial Intelligence - based automation tools for assessing digital pathology slides. This improves pathologists workflow and enables them to make more accurate diagnoses more efficiently.

I helped MindPeak achieve IEC 62304 - compliance by working closely with their CIO. I also helped set up their ISO 14971 - compliant risk analysis by explaning how the requirements of the standard can be applied to AI-based software.

Let's talk?

Send me an email
Digital Health Jobs No Cookie For You Privacy Policy Imprint
No QMS on this planet will save you from creating crappy software.