AI-300T00-A: Operationalize machine learning and generative AI solutions

Create and Configure Azure Machine Learning Workspaces

Manage Datastores, Compute Targets and Workspace Resources

Configure Identity and Access Management (IAM)

Create and Manage Data Assets, Environments and Components

Share Machine Learning Assets Using Registries

Implement Infrastructure as Code (IaC) with Bicep and Azure CLI

Configure GitHub Integration and Source Control

Automate Infrastructure Provisioning Using GitHub Actions

Implement Network Security and Access Restrictions

Manage Git-Based Machine Learning Projects

Configure Experiment Tracking with MLflow

Develop and Execute Model Training Scripts

Implement Automated Machine Learning (AutoML)

Manage Machine Learning Experiments Using Notebooks

Configure Hyperparameter Tuning and Optimization

Implement Distributed Model Training

Create and Automate Machine Learning Training Pipelines

Compare Model Performance Across Training Jobs

Register and Version Machine Learning Models

Evaluate Models Using Responsible AI Principles

Manage Model Lifecycle and Archiving

Deploy Models to Real-Time and Batch Endpoints

Test, Monitor and Troubleshoot Model Deployments

Implement Progressive Rollouts and Rollback Strategies

Detect Data Drift and Monitor Model Performance

Configure Automated Retraining and Monitoring Alerts

Create and Configure Microsoft Foundry Resources and Projects

Implement Identity and Access Management with Managed Identities

Configure Role-Based Access Control (RBAC)

Secure Generative AI Infrastructure Using Private Networking

Deploy Resources Using Bicep Templates and Azure CLI

Deploy Foundation Models for Production Workloads

Configure Serverless API Endpoints and Managed Compute

Select Appropriate Models for Enterprise AI Applications

Implement Model Versioning and Deployment Strategies

Configure Provisioned Throughput for High-Volume Workloads

Design and Manage Prompt Templates and Variants

Implement Prompt Versioning Using Git Repositories

Compare and Evaluate Prompt Performance

Design Evaluation Strategies for Generative AI Applications and Agents

Create and Manage Test Datasets

Configure Model Evaluation and Data Mapping

Evaluate Groundedness, Relevance, Coherence and Fluency

Implement Responsible AI and Content Safety Evaluations

Configure Automated Evaluation Workflows

Use Built-In and Custom Evaluation Metrics

Implement Continuous Monitoring in Microsoft Foundry

Monitor Latency, Throughput and Response Times

Track Token Consumption and Resource Usage

Analyze AI Operational Costs

Configure Application Logging and Distributed Tracing

Troubleshoot Production AI Applications and Agents

Monitor Generative AI Reliability and Quality

Optimize Retrieval-Augmented Generation (RAG) Solutions

Configure and Tune Retrieval Strategies

Optimize Similarity Thresholds and Document Chunk Sizes

Select and Fine-Tune Embedding Models

Implement Hybrid Search Combining Semantic and Keyword Retrieval

Evaluate RAG Relevance and Accuracy

Implement A/B Testing for Generative AI Systems

Optimize AI Application Response Quality and Performance

Design and Implement Advanced Model Fine-Tuning

Create and Manage Synthetic Training Data

Monitor Fine-Tuned Model Performance

Optimize Model Efficiency and Reliability

Manage Fine-Tuned Models Across Development and Production

Implement Model Customization and Deployment Strategies