Staff Engineer - Machine Learning

About the role

Western Digital is building the infrastructure behind the AI-driven data economy, combining engineering expertise with global-scale manufacturing to power hyperscale data centers, cloud platforms and enterprise infrastructure. This is a focused AI team solving hard scientific problems in precision product development, with mentorship from engineers and researchers in their first year. The role suits someone who learns fast and wants to be in the middle of difficult problems early in their career.

Key responsibilities

• Own MLflow experiment logging for assigned model runs, run model evaluations with standard metrics and produce structured reports that inform model iteration.
• Validate training datasets with the team: feature distribution checks, label verification and anomaly flagging, closing the data quality loop before data enters the training pipeline.
• Build and train CNN-based models for image classification and defect detection, and contribute to evaluation cycles, configuration comparisons and training run analysis.
• Package models in Docker, contribute to CI/CD scripts, run inference tests and support deployment validation in product development environments.

Required profile

• Bachelor's or Master's degree in AI, Machine Learning, Computer Science, Electrical Engineering, Applied Mathematics or a related field; AI major or strong research focus preferred.
• Fresh graduate to 1 year of experience; work experience not required, with demonstrated ML project competency being the main criterion.
• Strong final-year project or thesis with a clear ML component; research internship preferred.
• Self-directed learning and quick uptake of new concepts.

Required skills

• Python with clean, readable ML code and basic NumPy/Pandas.
• Foundational PyTorch: build, train and evaluate a basic neural network from scratch.
• CNN architecture basics and concept of data-efficient ML in limited-data scientific settings.
• Basic MLflow usage, Docker basics, and model evaluation reporting.
• Good to have: U-Net or ViT, uncertainty quantification, Bayesian methods, time-series or sensor data projects, RL basics, RAG basics, AWS fundamentals.