We are seeking a highly motivated AI Scientist specializing in ML-based Spatio-Temporal Prediction to join our growing AI R&D team. In this role, you will be at the forefront of developing and deploying cutting-edge ML models to solve real-world tabular, time series, and multimodal modeling challenges in manufacturing. We’re looking for a candidate with strong practical R&D experience, grounded in solid theoretical fundamentals, and deep expertise in AI disciplines. The ideal candidate will have a deep understanding of state-of-the-art machine learning algorithms and techniques, a track record of impactful publications in top-tier conferences such as NeurIPS, ICML, ICLR, KDD, CVPR, or ICCV, and a solid background in computer science and engineering. Experience collaborating with software engineering teams to scale and productize ML solutions is a strong plus. This is a high-impact role that combines foundational research, system-level design, and hands-on implementation. You’ll work closely with cross-functional teams to develop innovative solutions that guide strategic decisions and deliver tangible business value.
Responsibilities
- Develop and advance a Virtual Metrology (VM) algorithm and innovative prediction AI solutions using ML-based spatio-temporal prediction technology for semiconductor manufacturing processes
- Increase VM prediction accuracy based on multimodal and multidimensional data, including tabular, time series, and images generated in manufacturing process
- Build predictive and anomaly detection models based on diverse sensor and process data
- Analyze data and develop models using statistical, machine learning, and deep learning approaches
- Conduct feature engineering and variable optimization tailored to each manufacturing process
- Evaluate and optimize model performance, hyperparameter tuning, ML/DL automation (MLOps), and continual improvements
- Integrate model inference with manufacturing systems for both batch and real-time applications
- Ensure explainable AI (XAI) techniques such as SHAP for model transparency and interpretability
- Collaborate with process/quality/manufacturing engineers to design and optimize practical models for field deployment
Key Qualifications
- Ph.D. or Master’s degree with 5+ years of experience in Computer Science, Machine Learning, Statistics, or a related field.
- Proven experience in AI, machine learning, or deep learning development and application
- Experience with tabular data, time series, multivariate data analysis, anomaly detection, and predictive modeling
- Hands-on project experience with Python-based data analysis and machine learning (e.g., Scikit-learn, Pandas)
- Proficiency in deep learning frameworks such as PyTorch
- Familiarity with various AI algorithms such as regression, classification, decision trees, ensemble models (XGBoost, Random Forest), DNN, CNN, RNN, LSTM, Transformer, LLM, etc.
- Experience in feature engineering, model automation, hyperparameter optimization, and data preprocessing
- Strong publication track record in top-tier ML/AI conferences (e.g., NeurIPS, ICML, ICLR, CVPR, ICCV).
Preferred Qualifications
- Experience with data science in manufacturing industries (semiconductor, display, secondary battery, etc.)
- Understanding of manufacturing or semiconductor data and domain-specific characteristics is a plus
- Experience with integrating models with MES, SPC, FDC, or similar manufacturing/process systems is a plus.
- Record of publications or awards in AI/data science competitions
- Strong cross-functional communication and collaboration skills