MLOps & Model Deployment
Take ML models from notebook to production with CI/CD, monitoring, and retraining pipelines.
About this course
The engineering course for ML practitioners who are tired of models that only work in notebooks. You will build an end-to-end MLOps pipeline: experiment tracking with MLflow, model registry, containerisation with Docker, REST API serving, CI/CD with GitHub Actions, and a live monitoring dashboard with data-drift alerting.
Target audience: ML engineers, data scientists, DevOps engineers working on AI systems
What you will learn
- MLflow
- Docker
- CI/CD
- Model monitoring
- FastAPI
- Kubernetes basics
Course syllabus
10 modules · video + projects
- 1MLOps principles: reproducibility, versioning, and automation
- 2Experiment tracking with MLflow
- 3Data and model versioning with DVC
- 4Packaging models for production: Docker and environment management
- 5REST API serving with FastAPI and BentoML
- 6CI/CD for ML: automated testing and deployment with GitHub Actions
- 7Kubernetes basics for ML workloads
- 8Model monitoring: data drift, concept drift, and performance degradation
- 9A/B testing ML models in production
- 10Capstone: deploy a real-time fraud-detection service
Prerequisites
- –Python
- –Machine Learning Fundamentals
- –Basic Docker knowledge helpful
Frequently asked questions
Do I need cloud accounts?
The course uses free tiers of GitHub Actions, Docker Hub, and a small Railway deployment for the final project. Total cost is under £5 if free tiers are exhausted.
Ready to start MLOps & Model Deployment?
Join 2,400+ learners already enrolled. Self-paced, certificate on completion.