Job Description

Data Scientist (Medior)

About Refinery89

Refinery89 is a leading AdTech company helping digital publishers maximize the value of their advertising inventory through advanced programmatic monetization technology. Headquartered in Madrid, with a diverse international team, we combine data, automation, and innovation to make digital advertising more transparent, efficient, and impactful.

We are looking for an experienced and highly analytical Data Scientist (Medior) to join our growing AI and Data team. If you're passionate about building intelligent systems that drive real business impact, thrive on solving complex problems, and want to shape the future of AdTech innovation, we'd love to meet you.

What You'll Do

1. Lead the development of advanced statistical and machine learning models to optimize key performance parameters across our platforms (timeouts, refresh rates, ad placement logic, etc.).
2. Design and execute A/B/n experiments to test multiple parameter configurations simultaneously, driving measurable improvements in system efficiency and revenue outcomes.
3. Translate data and experimentation results into scalable, production-ready solutions that integrate seamlessly with our Product and Engineering teams.
4. Collaborate with cross-functional stakeholders—from Product to Commercial—to identify high-impact use cases and align technical outcomes with strategic goals.
5. Own the end-to-end lifecycle of models: from ideation, data wrangling, development, and deployment through to performance monitoring and optimization.
6. Present complex findings in a clear and actionable way to both technical and non-technical audiences.
7. Contribute to the continuous improvement of our data science practices, tooling, and experimentation frameworks.

Who We're Looking For

Experienced Problem Solver

• 2+ years of experience in Data Science, Machine Learning, or related analytical roles, ideally in fast-paced digital or AdTech environments.
• Strong analytical mindset with the ability to turn business challenges into data-driven solutions.

Advanced Technical Proficiency

• Strong Python and SQL skills.
• Proven experience with statistical modeling, experimentation, and machine learning frameworks such as scikit-learn, TensorFlow, or PyTorch.
• Solid understanding of data manipulation, feature engineering, and model evaluation techniques.

Cloud Native

• Familiarity with cloud environments such as AWS or Azure.
• Experience deploying and maintaining models within cloud infrastructure.

Experimentation Expert

• Hands-on experience designing, running, and analyzing A/B and multivariate tests with strong statistical rigor.
• Understanding of experimental design, hypothesis testing, and causal reasoning.

Communicator & Collaborator

• Able to communicate complex technical concepts clearly to both technical and non-technical stakeholders.
• Comfortable working across multiple teams and functions.

Business-Oriented Thinker

• Understands that the success of an algorithm is measured not only by accuracy but also by its impact on product performance and business outcomes.

Nice-to-Haves

• Background in AdTech, programmatic advertising, or digital publishing.
• Experience with LLMs, causal inference, or real-time prediction models.
• Exposure to tools such as Airflow, Docker, or Spark in production environments.

What You Can Expect from Us

• A fully remote working environment.
• A collaborative, international team with a strong knowledge-sharing culture.
• A flat and fast-moving organization inspired by our Dutch roots.
• Opportunities to work on high-impact projects at the intersection of AI, data, and programmatic advertising.
• Regular team gatherings, knowledge-sharing sessions, and company events.
• The opportunity to help shape the future of advertising technology through data-driven innovation.

Employment Details

• Permanent contract
• Full-time
• Remote (100%)

Ready to bring science to strategy and shape the future of advertising?

Apply today and help us shape the future of programmatic advertising.