For developers and Azure architects who need to deploy, secure and operate OpenAI and partner models and agents in Microsoft Foundry (formerly Azure AI Foundry).
The curriculum is being prepared and reviewed. Enrollment is not open yet.
About this course
Azure AI Foundry was renamed Microsoft Foundry in late 2025, and Azure OpenAI now lives inside it as one family of Foundry Models alongside models sold directly by Azure and partner models such as Anthropic's Claude. The portal, resource types, SDKs and API surface have all shifted: the v1 API lets you use the standard OpenAI SDK without dated api-version parameters, and Microsoft Agent Framework has replaced Semantic Kernel and AutoGen as the recommended agent SDK. This course uses the current names and points out the old ones you will still meet in tutorials and Terraform modules.
You will choose deployment types (Global, Data Zone or geography-based processing, each billed as Standard pay-per-token, Provisioned or Batch) against residency and throughput needs, call models with the Responses API using Microsoft Entra ID instead of keys, and configure content filters, Prompt Shields and groundedness detection. You will connect Azure AI Search for retrieval, decide when fine-tuning is justified, and build agents with Foundry Agent Service and Microsoft Agent Framework, including MCP tools. The final module covers TPM quotas and PTU sizing, private networking, Azure API Management as an AI gateway, and tracing and evaluations. Product facts are as of October 2026 and checked against Microsoft Learn.
What you will be able to do
Map the Microsoft Foundry resource and project model and explain where Azure OpenAI fits within Foundry Models
Select a deployment type that satisfies data-residency and throughput requirements and justify the choice
Call Foundry Models through the v1 API with the OpenAI SDK using Microsoft Entra ID and managed identity
Configure content filters, Prompt Shields and groundedness detection and interpret their annotations
Build a retrieval-augmented application on Azure AI Search and decide when fine-tuning is warranted
Build and deploy an agent with Foundry Agent Service or Microsoft Agent Framework that uses MCP tools
Size quota and provisioned throughput, and place Azure API Management and private endpoints in front of model deployments
Syllabus
6 modules · 17 lessons · about 12 hours
1
Module 1: The Foundry Platform Map
Resources, projects, model families and deployment types.3 lessons
From Azure AI Foundry to Microsoft Foundry: Resources, Projects and the Portal
Foundry Models: Azure OpenAI, Models Sold Directly by Azure and Partner Models
Deployment Types: Global, Data Zone or Geography Processing with Standard, Provisioned or Batch Billing
2
Module 2: Calling Models the Current Way
The v1 API, Responses and keyless authentication.3 lessons
The v1 API with the OpenAI SDK: Retiring Dated api-version Strings
Responses API, Function Calling and Structured Outputs on Azure
Microsoft Entra ID, Managed Identity and Keyless Access
3
Module 3: Content Safety and Data Handling
What Azure filters, what it logs and what you can change.3 lessons
Content Filters, Severity Thresholds and Custom Blocklists
Prompt Shields, Groundedness Detection and Protected Material Detection
Abuse Monitoring, Data Retention and Modified-Monitoring Requests
4
Module 4: Retrieval and Customization
Ground models in enterprise content and decide when to customize them.2 lessons
Azure AI Search as a Knowledge Source: Hybrid Retrieval and Semantic Ranking
Fine-Tuning in Foundry: When It Beats Prompting and How to Evaluate It
5
Module 5: Agents on Azure
Managed agents, the open-source agent SDK and tool connections.3 lessons
Foundry Agent Service: Tools, Conversations and Hosted Agents
Microsoft Agent Framework: The Successor to Semantic Kernel and AutoGen
Connecting Agents to MCP Servers and Publishing to Microsoft 365
6
Module 6: Production Operations
Capacity, network isolation, gateways and observability.3 lessons
Quotas, Tokens per Minute, Priority and Flex Processing, and PTU Sizing
Private Endpoints and Azure API Management as an AI Gateway
Tracing, Evaluations and Monitoring in Foundry
Exam and certificate
The exam unlocks once you have completed every lesson. It draws 20 questions from a larger bank, spread across the modules, with the answer options shuffled for each attempt. You have 40 minutes and need 80% to pass. After a failed attempt you can try again after 24 hours, up to 3 attempts in total. Feedback is given per module, never as the answer key.
Passing earns a certificate of completion and assessment from the Wirewalk Institute of AI with a unique ID that anyone can check at institute.wirewalk.com/verify. It is not an accredited qualification or a vendor certification. How certification works.
The Wirewalk Institute of AI is independent. This course teaches publicly documented products and is not affiliated with or endorsed by Microsoft (Foundry, Azure OpenAI, Azure AI Search, Agent Framework, Entra ID, API Management), OpenAI, Anthropic, Meta (Llama), Mistral AI, or DeepSeek. Product and company names are trademarks of their respective owners.
A hands-on hour with ChatGPT, Claude and Gemini, told through one small business racing a deadline. Set up an account with a privacy setting you chose, decide what never goes into a personal account, write prompts that produce usable drafts, use files and web search, catch confident mistakes before they go out, and pick the assistant that fits how you work.
A practical, current tour of ChatGPT as a professional tool: choosing a plan and model, prompting well, analyzing files, researching with citations, using images and voice, organizing work in projects and memory, connecting apps and plugins, delegating multi-step work to ChatGPT Work, and knowing which admin and privacy controls matter. Ends with clear guidance on moving a workflow to the OpenAI API.
Learn Claude by solving one firm's real problem: getting through renewal season with too few analysts and without putting client data in the wrong place. Each module compares the approaches professionals actually choose between, shows a realistic failure and its fix, and ends with a Decision Lab where you make the call under pressure.