Azure AI Foundry
I explore the capabilities of Azure AI Foundry for building and deploying intelligent agents.
A topic page: a summary of what my posts say, with each claim quoted from the post it came from. The posts are the authority. Last checked 2026-11-06.
Summary. I explore the capabilities of Azure AI Foundry for building and deploying intelligent agents. My posts demonstrate how to use the platform for complex tasks like legal analysis, voice interactions, and automated workflows.
What the posts say
I recommend keeping core logic in open layers to ensure portability.
“So I would put the valuable logic in the open layer, and let Foundry carry the plumbing.” — Where to start in Microsoft Foundry
The A2A tool presents remote agents as tools to the model.
“But to the model in the middle, I present it as a tool.” — Where to start in Microsoft Foundry
The supervisor design is the better choice if documents are too large for one prompt or if cost is a primary concern.
“If the documents did not fit in one prompt, or if the cost mattered more than the last six answers, the supervisor would be the answer.” — Azure Foundry Agent (Part 4)
The agent successfully retrieved and cited specific page and line numbers.
“I checked every citation against the transcript pages, and they were right.” — Azure Foundry Agent (Part 2)
The strict agent correctly used the knowledge base and provided clickable sources.
“I asked about DWI while asleep in a parked car, and the strict agent answered with the section name and two clickable references into the search index” — Azure Foundry Agent (Part 5)
The agent successfully cited the Benefit_Options.pdf document as the source for its answer.
“The agent cited Benefit_Options.pdf, and its answer was that health insurance costs are deducted from each paycheck.” — Azure Search (Part 2)
The Azure Functions MCP extension simplifies the process of turning a Python function into a tool.
“I chose an Azure Function, because the Functions MCP extension turns a Python function into an MCP tool with one decorator.” — Azure Foundry Agent (Part 3)
I used the Parse JSON action to extract specific values from the response.
“I used Parse JSON to extract the category, score, confidence and explaination values” — Power Automate
The posts behind this page
- Azure Foundry Agent (Part 7) — 2,000 simulated conversations for the Minnesota DWI advisor, six of them told as fireside chats, and the traps that GPT with search, GPT-5.5 and DeepSeek fall into that a rules base built from the statutes avoids
- Azure Foundry Agent (Part 6) — Rules from the statutes, a small model that only reads, and a side-by-side test against GPT, DeepSeek and Google on the same arrests
- Where to start in Microsoft Foundry — A 10,000-foot view of Microsoft Foundry for agents: what is open and cross-platform, what Microsoft adds and why, a suggested order to learn it in, and what is going away
- Azure Foundry Agent (Part 5) — Following Microsoft's Foundry IQ learning module with my six Minnesota law documents: a knowledge base, an agent with vague versus strict instructions, 17 questions, and a comparison with my other RAG posts
- Azure Foundry Agent (Part 3) — Starting from the Legal intake specialist voice agent template in Microsoft Foundry, adding a save tool as an MCP server on Azure Functions, and testing it with a made-up car accident call
- Azure Foundry Agent (Part 4) — Building a supervisor in Microsoft Foundry that calls six specialist agents over the Agent2Agent protocol, and testing whether the specialists notice a missing fact and ask for it, using the conversations from my LangChain post
- Azure Foundry Agent (Part 1) — Running my page-line-summary skill in Azure: I upload the skill to Microsoft Foundry, wrap it in a prompt agent, run it on deposition excerpts, and measure how reliably it finds contradictions
- Azure Foundry Agent (Part 2) — Building a hosted agent in Microsoft Foundry that loads my page-line-summary skill, reads deposition PDFs, searches an Azure AI Search index of the split case documents, and saves its summaries
- Azure Search (Part 2) — Creating an AI agent in Azure AI Foundry backed by Azure AI Search with vector embeddings for retrieval-augmented generation (RAG)
- Power Automate — Connect Power Automate to Azure AI Foundry