MLOps Fundamentals

Target audience

Students with no prior experience in AI/ML.

Delivery

Hands-on sessions facilitated by the instructor.

Modules

Module 1: Introduction to MLOps

• Understanding the need for MLOps
• Differences between MLOps, DevOps, and DataOps
• Overview of the MLOps lifecycle
• Roles and responsibilities of an MLOps engineer

Module 2: Data Management and Preprocessing

• Data collection and validation techniques
• Data transformation and feature engineering
• Tools: Pandas, NumPy, and Scikit-learn
• Best practices for data versioning

Module 3: Model Development and Experimentation

• Overview of model training and evaluation
• Experiment tracking with MLflow
• Hyperparameter tuning basics
• Introduction to model registries

Module 4: Model Packaging and Deployment (2 hrs)

• Containerization with Docker
• Serving models using BentoML
• Deployment strategies: REST APIs and batch processing
• Introduction to cloud deployment options

Module 5: CI/CD for Machine Learning (2 hrs)

• Setting up CI/CD pipelines for ML projects
• Integration with GitHub Actions and Jenkins
• Automating testing and deployment processes
• Monitoring pipeline performance

Module 6: Monitoring and Maintenance

• Monitoring model performance in production
• Detecting and handling data drift
• Logging and alerting mechanisms
• Tools: Prometheus, Grafana, and ELK stack

Module 7: Capstone Project and Review (3.5 hrs)

• End-to-end MLOps project implementation
• Review of key concepts and best practices
• Q&A session and feedback
• Discussion on real-world MLOps scenarios

Enroll

Chat with us on WhatsApp: Enroll via WhatsApp