Microsoft Azure: develop end-to-end solutions (AZ-204)

AI-900: Microsoft Azure AI Fundamentals (1 day)

Module 1: artificial intelligence workloads and their characteristics

• Identify the characteristics of common AI workloads:
  • data monitoring and anomaly detection workloads;
  • content moderation and custom workloads;
  • computer vision workloads;
  • natural language processing workloads;
  • knowledge mining workloads;
  • document intelligence workloads;
  • generative AI workload characteristics.
• Identify the core concepts of responsible AI:
  • considerations for fairness;
  • considerations for greater reliability and safety;
  • considerations for privacy and data security;
  • considerations for inclusiveness;
  • considerations for AI solution transparency;
  • considerations for accountability in AI solutions.

Module 2: machine learning fundamentals on Azure

• Identify common machine learning techniques:
  • regression scenarios;
  • classification scenarios;
  • clustering scenarios;
  • deep learning techniques.
• Introduce the main machine learning concepts:
  • features and labels in a dataset;
  • using training and validation datasets.
• Explore Azure Machine Learning capabilities:
  • automated machine learning features;
  • compute and data services for data science and machine learning;
  • model management and deployment features in Azure Machine Learning.

Module 3: characteristics of computer vision workloads

• Identify the main solutions:
  • image classification solution characteristics;
  • object detection solution characteristics;
  • optical character recognition solution characteristics;
  • face detection solution characteristics.
• Introduce Azure tools and services:
  • Azure AI Vision service features;
  • Azure AI face detection service features;
  • Face service features;
  • Azure AI Video Indexer service features.

Module 4: characteristics of natural language processing (NLP) workloads

• Identify common NLP workload scenarios:
  • characteristics and uses of key phrase extraction;
  • characteristics and uses of entity recognition;
  • characteristics and uses of sentiment analysis;
  • characteristics and uses of language modelling;
  • characteristics and uses of speech recognition and synthesis;
  • characteristics and uses of translation.
• Introduce Azure tools and services:
  • Azure AI Language service features;
  • Azure AI Speech service features;
  • Azure AI translation service features.

Module 5: generative AI workload capabilities

• Identify the characteristics of generative AI solutions:
  • generative AI model characteristics;
  • common generative AI scenarios;
  • generative AI considerations.
• Identify Azure OpenAI service capabilities:
  • natural language generation;
  • code generation;
  • image generation.

Microsoft Azure AZ-204: develop end-to-end solutions (4 days)

Module 1: implement IaaS solutions

• Provision virtual machines (VMs).
• Configure, validate and deploy ARM templates.
• Create container images for solutions.
• Publish an image to Azure Container Registry.
• Run containers using Azure Container Instance.

Module 2: create Azure App Service web applications

• Create an Azure App Service web application.
• Enable diagnostic logging.
• Deploy code to a web application.
• Configure web application settings, including SSL, API settings and connection strings.
• Implement autoscaling rules, including scheduled autoscaling and autoscaling based on operational or system metrics.

Module 3: implement Azure Functions

• Create and deploy Azure Functions applications.
• Implement input and output bindings for a function.
• Implement function triggers using data operations, timers and webhooks.
• Implement Azure Durable Functions.

Module 4: implement Azure Cosmos DB storage solutions

• Select the appropriate API and SDK for a solution.
• Implement partitioning schemes and partition keys.
• Perform operations on Azure Cosmos DB data and containers.
• Set the appropriate consistency level for operations.
• Manage change feed notifications.

Module 5: implement solutions requiring block storage

• Move items in block storage between storage accounts or containers.
• Set and retrieve properties and metadata.
• Perform data operations using the appropriate SDK and implement data storage, archiving and retention policies.

Module 6: configure user authentication and authorisation

• Authenticate and authorise users with the Microsoft Identity platform.
• Authenticate and authorise users and applications with Microsoft Azure Active Directory (Azure AD).
• Create and deploy shared access signatures.
• Implement solutions that interact with Microsoft Graph.

Module 7: implement secure cloud solutions

• Secure application configuration data using App Configuration or Azure Key Vault.
• Develop code that uses keys, secrets and certificates stored in Azure Key Vault.
• Implement managed identities for Azure resources.

Module 8: manage caching for Azure solutions

• Configure cache and expiration policies for Azure Cache for Redis.
• Implement secure, optimised application caching patterns, including data size, connections, encryption and expiration.

Module 9: troubleshoot solutions

• Configure an application or service to use Application Insights.
• Review and analyse metrics and log data.
• Implement Application Insights web tests and alerts.

Module 10: implement API management

• Create an APIM instance.
• Create and document APIs.
• Configure authentication for APIs.
• Define API policies.

Module 11: design event-driven solutions

• Implement solutions using Azure Event Grid.
• Implement solutions using Azure Event Hub.

Module 12: design message-driven solutions

• Implement solutions using Azure Service Bus.
• Implement solutions using Azure Queue Storage queues.