The course will cover the entire data engineering workflow, with emphasis on operations from a data and machine learning perspective (MLOps):

• Introduction to machine learning in production, machine learning lifecycle and deployment.
• Data ingestion and storage. Labelling and validating data, feature engineering, transformation, augmentation, and selection.
• Continuous integration and continuous deployment.
• Workflow orchestration.
• Data analysis reproducibility.
• Data analysis traceability (code, data, and model versioning).
• Continuous model training and evaluation. High-performance modeling and interpretability.
• Model serving: patterns and infrastructure, model management and delivery.
• Metadata tracking and logging.
• Monitoring. A/B testing.
• Data governance, security, privacy. FAIR principles.
• Integration into business.
• Real-world use cases overview.

The course is project-based. Students will gradually design, build, deploy, and monitor a solution with a data acquisition, data analysis, and reporting component to support end-to-end MLOps lifecycle.

Lectures will introduce the students to the principles and lectures from industry will introduce the students to the practical aspects of data engineering. Lab sessions will present the latest approaches and tools to support the students during their project work. Labs will also serve for consulting and evaluating project progress.

The set of tools used will be reviewed annually and updated to meet industry standards and developments.