Data Scientist
This role focuses on developing machine learning models for failure prediction and fault classification in semiconductor manufacturing equipment, leveraging time-series data and fault tree analysis. The candidate will engineer advanced features from sensor telemetry and logs to improve tool uptime and operational reliability.
Skills required
About this role
About the Role
We are seeking a Data Scientist with 5+ years of experience to develop machine learning solutions for failure prediction, classification, and fault analysis in semiconductor manufacturing and equipment systems. This role focuses on time-series modeling, equipment health monitoring, and root-cause analysis using structured reliability methods such as fault tree analysis (FTA).
You will work with complex, high-volume data from semiconductor tools (sensor signals, logs, process data) to improve tool uptime, yield, and operational reliability.
Key Responsibilities
- Design, develop, and deploy machine learning models for equipment failure prediction and fault classification
- Analyze time-series data from semiconductor tools (sensor telemetry, logs, process traces)
- Perform advanced feature engineering (lags, rolling windows, trends, seasonality, event-based features)
- Apply fault tree analysis (FTA) concepts to support root-cause analysis and improve model interpretability
- Collaborate with process engineers, equipment engineers, and failure analysis teams
- Select, justify, and evaluate appropriate ML algorithms
- Validate models using metrics such as precision/recall, F1-score, ROC-AUC, and early failure detection accuracy
- Document models, assumptions, and results for technical and cross-functional stakeholders
- Mentor junior data scientists and contribute to best practices
Work Experience
Required Qualifications
- 5+ years of professional experience as a Data Scientist or Machine Learning Engineer
- Strong proficiency in Python (Pandas, NumPy, scikit-learn)
- Proven experience with time-series data modeling
- Hands-on experience building classification and predictive models
- Experience with failure prediction, reliability analytics, or equipment health monitoring
- Working knowledge of fault tree analysis (FTA) or structured root-cause analysis
- Strong feature engineering skills for noisy, real-world industrial data
- Ability to clearly communicate technical results to engineering stakeholders
Preferred Qualifications
- Experience in semiconductor manufacturing or equipment systems (etch, deposition, lithography, inspection, metrology)
- Familiarity with process data, tool logs, alarms, and sensor telemetry
- Experience with survival analysis, RUL estimation, or anomaly detection
- Exposure to model explainability techniques (e.g., SHAP, feature importance)
- Experience deploying models into production or factory systems
- Background in reliability engineering, systems engineering, or failure analysis
What Success Looks Like
- Accurate and reliable failure prediction models with low false-positive rates
- Clear linkage between data-driven predictions and physical failure mechanisms
- Measurable improvements in tool uptime, yield, and maintenance planning
- Strong collaboration with cross-functional engineering teams
Representative Tech Stack
- Python (Pandas, NumPy, scikit-learn)
- Time-series analysis libraries
- Machine learning frameworks
- Visualization and reporting tools
