Medior Data Scientist

Job description

About Vitestro

Vitestro is pioneering the future of blood collection with the Aletta® Autonomous Robotic Phlebotomy Device™ (ARPD™). This medical device combines advanced multi-modal imaging (near-infrared, ultrasound, and Doppler ultrasound) with robotics and AI to perform the entire diagnostic blood draw procedure autonomously.

By addressing critical healthcare staffing shortages and improving patient experience, Vitestro is transforming one of the most common and essential medical procedures. With more than 90 team members and growing rapidly, the company is scaling its impact. As the first devices have been placed with customers, the team is expanding to ensure successful implementation and long-term reliability.

At Vitestro, we are committed to continuous innovation and improvement.

General Role Description

As our Medior Data Scientist, you play a key role in transforming complex device and procedural data into actionable insights that support product development, clinical validation, and operational improvements. You operate at the intersection of engineering, product, clinical, and data teams, developing probabilistic models and analytical frameworks that help guide decision-making across the organization. You actively collaborate with stakeholders to align on data quality, validation approaches, priorities, and scalable deployment of models.

First months at Vitestro

In your first months, your focus will be on understanding our device data, procedural workflows, and the KPIs that drive system performance. You will independently navigate large and imperfect datasets, align with stakeholders across engineering, product, clinical, and data teams, and establish effective ways of working for model development and deployment.

You will develop a probabilistic puncture model in a Jupyter Notebook environment to support hypothesis-driven testing of software improvements. Alongside this, you validate assumptions, improve data quality insights, and ensure analytical results are reproducible and clearly documented.

You will also strengthen analytical capabilities within the engineering organization by sharing practical data modeling techniques through hands-on sessions tailored to current analytical challenges.

This is what your success looks like

After 12 months, at least one of your models has been clinically validated through field testing and is actively used to support product and engineering decisions. Your models are accessible to - and trusted by - stakeholders across multiple disciplines and their documentation is of such quality that it allows independent usage. Moreover, they require only minimal maintenance effort.

You have also identified and validated an initial predictive maintenance signal using field data, resulting in at least one proactive maintenance action or design recommendation.

You thrive in a multidisciplinary environment where engineering, clinical, and product management teams collaborate closely. You enjoy translating complex and imperfect data into practical insights and bring a pragmatic, hands-on mentality to solving challenging problems.

Job requirements

What You’ll Bring:

• Strong proficiency in Python and the standard ML/data stack (such as scikit-learn, pandas, NumPy, and PyTorch or similar).
• A proven track record in statistics and probabilistic modeling, including topics such as Bayesian reasoning, hypothesis testing, probability distributions, and uncertainty quantification.
• Experience with causal inference, probabilistic graphical models, survival analysis, or similar methods for analyzing sequential or time-dependent processes.
• Comfort working with large, complex, and imperfect datasets — you know how to validate, clean, and reason critically about data quality.
• Strong communication skills and the ability to explain complex analytical topics to diverse technical and non-technical stakeholders.
• A proactive, hands-on, and pragmatic mindset with the ability to work independently in a multidisciplinary environment.
• Please note that you need to be authorized to work in the Netherlands — unfortunately, visa sponsorship is not available for this role.

Nice to have:

• Experience working in a regulated industry such as medical devices, pharma, aerospace, or other safety-critical environments.
• Familiarity with predictive maintenance techniques, such as anomaly detection or remaining useful life (RUL) estimation on time-series sensor data.
• Experience with process mining, Markov models, or workflow analysis techniques.
• Exposure to computer vision, multi-modal sensor fusion, or medical imaging data.
• Knowledge of probabilistic programming frameworks such as PyMC.
• Experience with AWS data services (such as S3, Glue, Athena, or SageMaker) and cloud-based data infrastructure.

What we offer

• Competitive salary including pension plan
• 25 days of annual leave based on a full-time position
• Hybrid work model
• Working together in a high-skilled team with our home base in Utrecht
• Frequent out-of-work activities with our team, and annual company weekend away.
• Training budget
• Opportunity to be a key shaper of a new global industry niche: autonomous medical robotics

We’re big fans of working from the office – this is where team spirit comes alive. At the same time, we offer flexibility, allowing up to two (out of 5) days of working from home for those who prefer it.

We offer an environment where you can make a lot of impact, as you are joining a growing scale-up in MedTech. Our team of driven and talented colleagues enjoys solving the challenges in developing, manufacturing, and completing our breakthrough medical device. While working in an open culture, we strive for excellence in all domains. We believe in taking courage and like to challenge each other. We learn and teach.