Description
MLOps: Machine Learning in Production is a written, self-paced course in six modules. You cover ML pipelines with orchestrators such as Kubeflow, Airflow and Prefect; experiment tracking with MLflow and Weights & Biases; model versioning, registries and CI/CD for ML; containerising ML workloads with Docker and Kubernetes; feature stores and data versioning with DVC; and monitoring model performance and data drift in production.
It is an educational course with practical exercises, not a vendor certification, and it is not affiliated with any tool mentioned. It assumes some Python and machine learning experience, and it does not promise a job outcome. Tools change quickly, so check current documentation when you set them up.
What you will be able to do
- Build a reproducible training pipeline with an orchestrator
- Track experiments, parameters and metrics so results can be compared
- Register, version and promote models through CI/CD
- Package a model as a container and deploy it to Kubernetes
- Monitor a deployed model for performance and data drift
Educational content only. No income, revenue or earnings results are promised or implied. This course does not award a certification, licence or professional credential.
