Senior Data Scientist – AI, ML, Predictive System Maintenance
Posted 2hrs ago
Employment Information
Report this job
Job expired or something wrong with this job?
Job Description
Senior Data Scientist developing machine-learning models for Capgemini Engineering’s industrial predictive-maintenance solutions. Analyzing telemetry and equipment data to enable proactive maintenance decisions.
Responsibilities:
- Analyze historical equipment, maintenance, operational, sensor and telemetry data to identify patterns associated with equipment degradation and failures
- Work with domain experts to understand failure modes, potential indicators, and whether available data provides sufficient predictive signals
- Perform exploratory data analysis, data profiling, statistical analysis, correlation analysis and feature discovery
- Assess data quality, completeness, frequency and predictive value before model development
- Identify and engineer relevant features from time-series and operational data
- Develop, train, validate and optimize machine learning models for failure prediction, anomaly detection and predictive maintenance
- Evaluate appropriate modeling approaches for different equipment and failure scenarios
- Define suitable model evaluation metrics and validate model performance against business and operational objectives
- Develop approaches for handling imbalanced datasets and rare failure events
- Collaborate with Data Engineers and solution architects on data preparation and model integration
- Collaborate with equipment and maintenance SMEs to validate model outputs and ensure predictions are operationally meaningful
- Support model explainability and provide transparency into the major factors contributing to predictions
- Document experiments, assumptions, data limitations, model performance and recommendations
- Contribute to transitioning validated models from experimentation or proof of concept into a production-ready solution
- Help determine which equipment failures can realistically be predicted from available data and establish relationships between telemetry signals and failure modes
- Build validated predictive models supporting proactive maintenance decisions
Requirements:
- 5+ years of experience in Data Science or Machine Learning, preferably including predictive maintenance, industrial analytics or IoT use cases
- Strong proficiency in Python and common data science libraries such as Pandas, NumPy and Scikit-learn
- Strong knowledge of machine learning, statistical analysis, time-series analysis, anomaly detection, classification, regression, feature engineering, and model validation and optimization
- Experience working with sensor, telemetry, event and time-series datasets
- Understanding of predictive maintenance concepts, including failure prediction, condition monitoring and equipment-health indicators
- Experience dealing with missing, noisy, sparse and imbalanced industrial data
- Strong SQL and data-analysis skills
- Experience with model explainability and communicating model results to technical and business stakeholders
- Ability to work closely with engineering and domain SMEs to translate physical equipment behavior into analytical hypotheses
- Previous experience with manufacturing, industrial equipment, IoT, vending or cooling equipment, or connected-device analytics
- Experience with cloud-based ML and data platforms
- Experience with Azure Machine Learning, Databricks, MLflow or equivalent technologies
- Knowledge of MLOps and production model monitoring
- Experience with streaming or near-real-time telemetry
- Understanding of Remaining Useful Life modeling and survival or reliability analysis
- Experience with deep-learning approaches for time-series data
Benefits:
- Health insurance from the first days, regardless of the probationary period
- Christmas holidays from 25 December to 31 December
- Cooperation with Superhumans center and Veteran HUB
- Psychological counseling provided by the Veteran Hub
- Internal policy making the company friendly to military and veterans


















