JOB TITLE: DATA SCIENTIST AI & MACHINE LEARNING - MEDIOR
LOCATION:63457 HANAU, GERMANY
LANGUAGES REQUIRED: ENGLISH, GERMAN
CONTRACT DURATION: ASAP - 31/12/2026
WORK MODE: FULL TIME (HYBRID AND 3 DAYS ONSITE PER WEEK)
JOB DESCRIPTION
Company Overview
- Client is a global leader in materials technology and recycling, headquartered in Belgium. The company focuses on clean mobility and circular economy solutions, developing high-performance materials for rechargeable batteries, automotive catalysts, and precious metals recycling. With a strong commitment to sustainability, Client aims to reduce harmful emissions, power the vehicles of the future, and give new life to used metals.
Department Description and Responsibilities
- The position is part of Client's Computational Science - Data Science team, operating within the Corporate Research & Development department. The team supports internal customers with data science, timeseries forecasting, machine learning, Generative AI, …
- The department works at the interface of research, product development, engineering, and business stakeholders. Its purpose is to translate scientific, experimental, product, and process data into reliable analytical models and practical decision support for Client's business units.
- Develop and apply a wide range of machine learning, statistical modelling, and predictive analytics methods to accelerate research and product development.
- Support internal stakeholders with predictive analytics, scientific computing, and data-driven decision-making.
- Collaborate with domain experts, engineers, and data engineering teams to transform complex technical data into actionable insights.
- Contribute to reusable modelling approaches, robust ways of working, and scalable analytical solutions across Corporate R&D.
Project Description
- This assignment focuses on developing predictive machine learning models that support product development and performance optimisation using historical experimental and operational data. The objective is to enable data-driven decision-making by predicting product behaviour and key performance indicators early in the development cycle, reducing reliance on trial-and-error experimentation.
- Project objective: Develop robust machine learning models capable of predicting product performance and supporting optimisation decisions across R&D and product development activities.
- Scope: Accelerate the end-to-end model development process, including data exploration, data quality assessment, dataset preparation, feature engineering, correlation analysis, and predictive modelling.
- Key challenge: Transform large, complex, and often imperfect datasets into reliable insights and predictive capabilities that create measurable business value.
- Stakeholders: Cross-functional collaboration with R&D, product development, data science, data engineering, and business teams.
- Business impact: Faster product development cycles, improved use of historical data, enhanced decision-making, and increased confidence in product optimisation strategies.
- Profile sought: A hands-on Data Scientist with experience in exploratory data analysis, data preparation, feature engineering, and predictive modelling. Comfortable working with complex datasets and translating business challenges into practical machine learning solutions.
Primary Tasks and Responsibilities
Core Responsibilities
- Develop, validate, and improve machine learning models for catalyst performance prediction.
- Build predictive models for full driving cycle emissions using scientific, experimental, product, and engineering datasets.
- Model the impact various product features and design parameters on performance outcomes.
- Translate business and scientific questions into data science approaches, modelling strategies, and measurable deliverables.
- Perform exploratory data analysis, feature engineering, model selection, hyperparameter tuning, and model performance assessment.
- Document model assumptions, methods, validation results, limitations, and recommendations in a clear and reproducible way.
Data Preparation and Correlation Analysis
- Explore relevant databases and available data sources to understand data structures, completeness, quality, and usability for modelling.
- Clean, transform, and structure raw or inconsistent datasets into reliable analytical datasets.
- Identify correlations, patterns, outliers, and potential performance drivers within catalyst, emissions, laboratory, product, and process data.
- Assess data quality issues and work with data engineering or business stakeholders to resolve gaps, inconsistencies, and traceability concerns.
- Create repeatable analysis workflows that support efficient re-use of cleaned datasets and feature sets.
Scientific and Industrial Data Responsibilities
- Work with catalyst, process, product, and laboratory data to identify performance drivers and relevant modelling features.
- Collaborate with data engineering resources to define data requirements, improve data quality, and create reliable modelling datasets.
- Ensure traceability of data sources, transformations, model versions, and analytical outcomes.
- Support the design of analytical datasets suitable for model training, validation, and future extension.
Project Delivery and Stakeholder Management
- Work closely with Automotive Catalysts stakeholders and Corporate R&D experts to align model outputs with project needs.
- Translate complex model results into practical recommendations for product design and development decisions.
- Contribute to accelerated project execution and help remove the data science bottleneck identified for Project Raptor.
- Present findings to technical and non-technical stakeholders in a structured and business-oriented manner.
- Support the transition from exploratory analysis and prototype models to robust, usable modelling assets for the project team.
Technical Profile Requirements
Mandatory Requirements
- Master's degree in Data Science, Statistics, Mathematics, Computer Science, Engineering, Physics, Chemistry, Materials Science, or a related discipline.
- Approximately 2-4 years of relevant experience in data science, machine learning, statistical modelling, or predictive analytics.
- Hands-on programming experience with Python and commonly used data science libraries such as pandas, NumPy, scikit-learn, matplotlib, or equivalent tools.
- Practical experience with database exploration, data cleaning, data preparation, exploratory data analysis, and correlation analysis.
- Experience developing, validating, and improving predictive models using complex technical or scientific datasets.
- Strong understanding of feature engineering, model evaluation, data quality, reproducibility, and interpretation of analytical results.
- Experience working with structured data from industrial, engineering, manufacturing, laboratory, or R&D environments.
- Ability to create clear visualisations, technical documentation, and model interpretation materials.
Preferred Requirements
- Experience in chemical engineering, catalyst development, emissions systems, materials science, automotive R&D, or related industrial research environments.
- Experience with supervised learning, ensemble methods, gradient boosting, regression/classification models, Bayesian modelling, hybrid modelling, or physics-informed machine learning.
- Experience working with SQL databases, data pipelines, large datasets, or industrial data platforms.
- Familiarity with Azure Machine Learning, Databricks, cloud-based data platforms, or comparable enterprise data science environments.
- Experience with experimental design, laboratory data, process data, or product development data.
- Strong ability to work with domain experts to connect scientific parameters and engineering constraints to machine learning features and model outputs.
Non-Technical Profile Requirements
- Communication skills: able to explain data quality findings, correlations, model outcomes, and limitations to technical and non-technical audiences.
- Leadership capabilities: able to take ownership of assigned analysis and modelling workstreams without requiring senior-level direction for every step.
- Stakeholder management skills: comfortable working with business stakeholders, catalyst experts, engineers, data engineers, and leadership representatives.
- Problem-solving capabilities: strong analytical mindset with the ability to connect data, domain knowledge, and business objectives.
- Hands-on mindset: willing to spend a significant part of the assignment on database exploration, data cleaning, dataset preparation, and correlation analysis.
- Collaboration: effective in cross-functional and international R&D environments.
- Delivery orientation: pragmatic, structured, and focused on creating usable datasets, insights, and models that support real product development decisions.
- Quality mindset: attentive to data quality, documentation, reproducibility, and responsible use of AI methods.
- Methodology: Experience working in Agile, Scrum, Kanban, or iterative R&D delivery environments is preferred.
- Training requirement: Continuous learning mindset in classical AI, machine learning, data science, data engineering basics, and industrial analytics is expected.
- Preferred training: MLOps, data engineering, scientific computing, statistics, or software engineering training is advantageous.
Skills
- 2-4 years of relevant experience in data science, machine learning, statistical modelling, or predictive analytics
- Ability to create clear visualisations, technical documentation, and model interpretation materials
- Experience developing, validating, and improving predictive models using complex technical or scientific datasets
- Experience working with structured data from industrial, engineering, manufacturing, laboratory, or R&D environments
- Hands-on programming experience with Python and commonly used data science libraries such as pandas, NumPy, scikit-learn, matplotlib, or equivalent tools
- Practical experience with database exploration, data cleaning, data preparation, exploratory data analysis, and correlation analysis
- Strong understanding of feature engineering, model evaluation, data quality, reproducibility, and interpretation of analytical results
#J-18808-Ljbffr
