Azure Machine Learning Engineering Training (DP-100)

Design and implement end-to-end machine learning solutions on Azure with this comprehensive 4-day training. Learn to build, train, and deploy models using Azure Machine Learning workspace, AutoML, ML pipelines, and MLOps practices while preparing for the Azure Data Scientist Associate (DP-100) certification.

Training Details


Duration	4 days (32 hours)
Level	Advanced
Delivery	In-person, Live online, Hybrid
Certification	Microsoft Certified: Azure Data Scientist Associate (DP-100)

Who Is This For?

Data scientists building ML solutions on Azure
ML engineers implementing production machine learning pipelines
Software engineers transitioning into machine learning roles
Data engineers working with ML model deployment and operations
Anyone preparing for the Azure Data Scientist Associate (DP-100) certification

Learning Outcomes

After completing this training, you’ll be able to:

Set up and manage Azure Machine Learning workspaces and compute resources
Build automated ML experiments for classification, regression, and forecasting
Train custom models using the Azure ML Python SDK v2
Design and orchestrate ML pipelines with reusable components
Deploy models to managed endpoints for real-time and batch inference
Implement MLOps practices with CI/CD and responsible AI dashboards

Detailed Agenda

Day 1: Azure ML Workspace and Data Management

Module 1: Azure Machine Learning Workspace and Compute Setup

Azure ML workspace architecture and components
Compute instances, compute clusters, and serverless compute
Environment management and custom Docker images
Azure ML CLI v2 and Python SDK v2 overview
Hands-on: Set up workspace and configure compute resources

Module 2: Data Assets and Datastores

Registering and versioning data assets
Connecting to Azure Blob Storage, Data Lake, and SQL databases
Data types: URI file, URI folder, and MLTable
Data exploration and profiling in Azure ML Studio
Hands-on: Register data assets and explore datasets

Module 3: AutoML for Classification, Regression, and Forecasting

Automated ML concepts and supported task types
Feature engineering and data preprocessing in AutoML
Model selection, hyperparameter search, and ensemble methods
Interpreting AutoML results and model explanations
Hands-on: Run AutoML experiments and evaluate results

Day 2: Custom Model Training and Pipelines

Module 4: Custom Model Training with SDK v2

Training scripts and command jobs
Distributed training across compute clusters
Experiment tracking with MLflow integration
Logging metrics, artifacts, and model outputs
Hands-on: Train a custom model with experiment tracking

Module 5: ML Pipelines and Components

Pipeline architecture and design patterns
Building reusable components for data prep, training, and evaluation
Pipeline scheduling and triggering
Parameterized pipelines for experimentation
Hands-on: Build an end-to-end ML pipeline

Module 6: Hyperparameter Tuning

Sweep jobs and hyperparameter search spaces
Sampling methods: grid, random, and Bayesian
Early termination policies for efficient tuning
Analyzing sweep job results
Hands-on: Optimize model performance with hyperparameter sweeps

Day 3: Model Deployment and Batch Inference

Module 7: Model Registration and Deployment

Model registration and versioning in the model registry
Managed online endpoints for real-time inference
Scoring scripts and custom inference configurations
Blue-green deployment and traffic splitting
Hands-on: Deploy a model to a managed online endpoint

Module 8: Batch Inference

Batch endpoints for large-scale predictions
Parallel run step for distributed batch processing
Scheduling batch inference jobs
Output management and result aggregation
Hands-on: Set up batch inference pipeline

Day 4: MLOps, Responsible AI, and Production Operations

Module 9: MLOps with GitHub Actions

MLOps principles and maturity model
CI/CD pipelines for ML with GitHub Actions
Automating model training, testing, and deployment
Infrastructure as code for Azure ML resources
Hands-on: Build a CI/CD pipeline for model deployment

Module 10: Responsible AI Dashboard

Responsible AI concepts and pillars
Fairness assessment and bias detection
Model interpretability and explainability
Error analysis and counterfactual explanations
Hands-on: Generate and interpret a responsible AI dashboard

Module 11: Model Monitoring in Production

Data drift detection and monitoring
Model performance degradation alerts
Feature importance tracking over time
Retraining triggers and automation
Hands-on: Set up model monitoring and drift detection

Module 12: Exam Preparation

DP-100 exam format and question types
Practice scenarios and case studies
Study resources and exam tips

Prerequisites

Proficiency in Python programming
Understanding of machine learning fundamentals (supervised/unsupervised learning, model evaluation)
Basic familiarity with Azure portal and services
Experience with data manipulation libraries (pandas, NumPy, scikit-learn)

Delivery Formats

Format	Description
In-Person	On-site at your company’s location, hands-on with direct interaction
Live Online	Interactive virtual sessions with screen sharing and real-time labs
Hybrid	Combination of on-site and remote sessions, flexible scheduling

All formats include hands-on labs, course materials, practice exams, code samples, and post-training support.