The Predictive Models team inside the Marketing and Communication department of a major Belgian financial institution builds responsible AI models to make marketing campaigns more relevant and measurable. This senior data scientist role focuses on building production-ready machine learning and deep learning solutions using Python and LLMs for text classification and summarization, and on integrating those models into customer-journey driven campaign pipelines.
The mission
The team develops models that identify relevant target groups and capture customer intent from web analytics and text sources, applying statistical methods and experiment design to measure campaign impact. Work touches campaign tooling used by marketing teams, with models feeding targeting and measurement workflows and following the organisation's responsible AI guidelines for explainability and validation.
On a day-to-day basis you will take business requirements through feasibility, modelling and deployment: cleaning and structuring analytics data, training traditional and deep learning models, applying LLMs for text classification and summarization, and setting up A/B and uplift tests to quantify lift. You will collaborate with data engineers, product owners and marketing stakeholders to move models from prototype to production and maintain monitoring and validation pipelines.
Your responsibilities
- Design and deliver production machine learning models that increase campaign relevance and measurable lift, prioritising explainability and validation.
- Build and maintain data pipelines that transform web analytics and CRM data into modelling-ready datasets and customer journey features.
- Implement and validate LLM-based text classification and summarization components, including guardrails and performance monitoring.
- Define and run experiment design and uplift tests to quantify model impact and inform campaign decisions.
- Deploy and monitor models in production, automate retraining or incremental learning where appropriate, and keep stakeholders informed with clear technical-to-business translations.
- Advise marketing teams on how to use model outputs responsibly in campaign targeting and measurement.
Your profile
Essential skills
- At least 5 years' experience building and shipping machine learning models in production.
- Strong Python programming skills and fluency with the Python data science stack, including scikit-learn, PyTorch or TensorFlow.
- Practical experience with LLMs for text classification and summarization, plus methods for interpretability, guardrails and validation of LLM-driven predictions.
- Solid knowledge of statistical methods, experiment design and traditional ML algorithms.
- Proficient with SQL, git and bash for data access, versioning and automation.
- Ability to translate technical results into actionable business recommendations and to run projects end-to-end with limited supervision.
Preferred skills
- Experience with incremental learning, uplift modelling or Marketing Mix Modeling.
- Familiarity with SAS and Streamlit for lightweight tooling and prototypes.
Languages
- English, C1
- Dutch, C1 or French, C1
Education
- Masters degree or higher in a quantitative field such as statistics, mathematics or computer science, or equivalent practical experience.