COURSE & TRAINING
Effective MLOps
Operationalize machine learning models with reproducible workflows, deployment pipelines, testing, monitoring, and responsible AI practices.
Online
CHF 1'250
Live, instructor-led public or private delivery.
On-site
CHF 2'500
On-site delivery for team or public cohorts.
Private team cohort
On request
Private delivery tailored for your engineering team.
Who this training is for
Data Scientists
Move models from notebooks into reproducible, versioned, and testable delivery workflows.
ML Engineers
Build deployment, monitoring, and validation patterns for production machine learning systems.
DevOps Engineers interested in production ML
Adapt CI/CD, containers, and operational practices to the ML lifecycle.
What you'll learn
Build reproducible ML pipelines
Implement tracked and versioned machine learning workflows with reproducible experiments and data pipelines.
Deploy ML systems end to end
Move from notebooks to production code, package services, and deploy full ML applications with CI/CD workflows.
Test and monitor ML in production
Apply unit and integration testing, validation checks, drift monitoring, and production monitoring patterns.
Understand governance and responsible AI
Address explainability, bias, and operational governance concerns as part of the ML delivery lifecycle.
PREREQUISITES
What participants should bring
Participants should have proficiency in Python, a basic understanding of machine learning algorithms and model development, and familiarity with Git and command-line tools. Knowledge of Docker and cloud platforms is a plus.
PREPARATION
Before the course starts
Participants should be ready for a practical case study involving experiment tracking, deployment, testing, and monitoring across a full ML pipeline.
CURRICULUM
Inside Effective MLOps
This course is available online, onsite, and as a private team cohort. Public sessions proceed subject to minimum participant numbers and can be adapted for internal ML platforms and governance requirements.
Foundations of MLOps and Experiment Management
Day 1
The scope of MLOps, data exploration, and tracked experimentation with versioned models.
Foundations of MLOps and Experiment Management
Day 1
The scope of MLOps, data exploration, and tracked experimentation with versioned models.
Theoretical Part
- Definition and scope of MLOps
- Comparison with DevOps
- ML lifecycle overview
- Business value and challenges
- MLOps maturity levels
Practical Part
- Set up the environment with Git, Docker, and Python tooling
- Explore the fraud detection use case and dataset
- Handle missing values, outliers, and class imbalance
- Apply feature engineering techniques
- Train, version, and track experiments with DVC and MLflow
Modularization, Deployment, and CI/CD Pipelines
Day 2
Turn notebooks into deployable systems and build a full ML application delivery pipeline.
Modularization, Deployment, and CI/CD Pipelines
Day 2
Turn notebooks into deployable systems and build a full ML application delivery pipeline.
Theoretical Part
- Principles of clean ML code
- Converting notebooks into modular scripts
- Packaging ML code into reusable components
- Web application structure with FastAPI and Streamlit
- Containerization and cloud deployment concepts
Practical Part
- Refactor notebooks into production-oriented Python modules
- Build a REST API with FastAPI and a frontend with Streamlit
- Dockerize backend and frontend services
- Use Docker Compose for local deployment
- Work through AWS ECS, ECR, and GitHub Actions deployment steps
Testing, Monitoring, and Responsible AI
Day 3
Validation, production monitoring, and governance patterns for operational ML.
Testing, Monitoring, and Responsible AI
Day 3
Validation, production monitoring, and governance patterns for operational ML.
Theoretical Part
- Unit testing for ML components
- Integration testing of pipelines
- Data and model validation with Deepchecks
- Model and data drift concepts
- Responsible AI including bias detection and explainability
Practical Part
- Write and organize ML system tests
- Perform validation checks for data and model behavior
- Build and review a complete MLOps pipeline
- Integrate DVC, MLflow, FastAPI, Streamlit, Docker, and GitHub Actions
- Present and review the final solution
TECHNOLOGIES
Technologies used in Effective MLOps
These are the tools used in the public version of the course. For private cohorts, we can swap the stack and tune the flow to your team.
Need Effective MLOps for your team?
Private cohorts can be run online or on-site, adapted to your tool stack, and reshaped into a half-day, multi-day, or mixed-course program for your team.
COURSE INSTRUCTORS
Karim Baklouti
Instructor - ML & Data Platforms
Sonia Boussabeh
Instructor - Data & ML Engineering
Upcoming cohorts
Date / PeriodSession 1: June 2026Online — Online via Teams
StatusOpen
Actions
Frequently asked questions
Everything you need to know about Effective MLOps.
Who is Effective MLOps designed for?
The course is aimed at ML engineers, data science teams, and platform teams working to productionize machine learning systems with better delivery, testing, and monitoring practices.
Does the MLOps course cover the full model lifecycle?
Yes. It covers experiment tracking, versioning, application modularization, deployment, CI/CD, testing, monitoring, and responsible AI concerns.
Can Effective MLOps be delivered privately for a team?
Yes. The course can be adapted for internal ML platforms, model governance requirements, and team-specific tooling or cloud infrastructure.