Microsoft Foundry Solution Accelerators by Industry

The question comes up in nearly every Microsoft Foundry conversation: is there an official example for my industry? The answer is more useful than a plain yes or no, and it is not on one page, so this post lays it out.

Microsoft lists its Foundry-era solution accelerators at accelerators.ms. The page is rendered in the browser, so the catalog data is easier to read in the Solution-Accelerators repository that powers it. Everything below was checked against that data and each accelerator's own repository on September 26, 2026.

How the catalog is organized

There are 15 accelerators, all labeled Live, all MIT-licensed, none archived. Each is built around a functional pattern, not an industry: chat over your own data, content processing, knowledge mining, multi-agent automation, code modernization, data governance, and so on.

Industry appears in three weaker ways, and the difference matters:

• A dedicated walkthrough deck for an industry, attached to a functional accelerator
• A named use case in the accelerator's description
• A scenario pack inside the repository, such as sample data and agent instructions for one industry

None of these is a separate, production-ready industry product. They are starting points, and a few are stronger starting points than others.

What exists, by industry

Industry	Best official starting point	How strong the match is
Energy	Chat with your data, Energy walkthrough deck; Microsoft IQ	Dedicated deck; named use case ("unified data and analytics agents for energy")
Insurance	Agentic Applications for Unified Data Foundation; Content Processing	Dedicated Insurance deck; named use case (claims processing)
Healthcare	Content Processing; Customer Chatbot; Prior Authorization Multi-Agent (outside the catalog)	Named use case (clinician-patient visit documentation); scenario pack on sample data; a payer-side prior authorization reference
Banking and financial services	Chat with your data, Financial Advisor deck; Customer Chatbot	Dedicated deck; scenario pack on sample data; Document Knowledge Mining carries a "banking" tag only
Retail	Customer Chatbot; Agentic Applications for Unified Data Foundation; Microsoft IQ	Scenario pack; named use cases (product discovery, retail supply chain)
Telecom	Conversation Knowledge Mining, Telco deck	Dedicated deck
Manufacturing and industrials	Real-Time Intelligence for Operations; Microsoft IQ; Data and Agent Governance and Security	Named use cases; no dedicated deck
Life sciences	Document Knowledge Mining	Named use cases (drug discovery, research)
Automotive	Microsoft IQ	One named use case

Three notes on reading that table honestly. First, "dedicated deck" means a Microsoft-authored presentation, not a separate repository. Second, the healthcare and banking scenarios in the chatbot accelerator run on fictional sample companies, so they demonstrate the pattern rather than a regulated product. Third, the prior authorization accelerator is a separate repository, not one of the 15 catalog entries. It is active and MIT-licensed, built as a payer-side review flow with multiple hosted agents and human-in-the-loop oversight, and it is the clearest example of an accelerator built for one regulated vertical.

The horizontal accelerators

Several accelerators have no industry framing at all and are useful across all of them:

• Multi-Agent Custom Automation Engine, with RFP reviewer and contract compliance reviewer walkthroughs
• Modernize your Code and Container Migration, for code and platform migration
• Multi-Agent Content Generation, for marketing content
• Deploy Your AI Application in Production, for a secure production environment
• Data and Agent Governance and Security, for Purview-based governance and compliance guardrails
• Unified Data Foundation with Fabric, the data layer several others build on

What to check before you build on one

Treat an accelerator as a fast way to see a pattern working, not as a shortcut past design decisions. Before committing:

• Read the accelerator's own repository, not just its catalog card. Status, last commit, and open issues tell you more than the summary.
• Replace the sample data early. A scenario that works on fictional data proves the wiring, not the fit.
• Run it on a governed platform. Every accelerator here deploys onto infrastructure someone has to secure. The same identity, networking, and policy decisions from Azure AI Landing Zones apply regardless of which accelerator you pick, and running agentic AI at enterprise scale explains why they have to come first.
• Do not skip a compliance review. An industry deck is not one.

If you would rather learn the platform hands-on first, our open-source Introduction to Microsoft Foundry workshop takes about an hour.

For a worked example of choosing a starting point and putting a governed platform under it, see how we approached standing up a governed agentic AI platform for an oil and gas environment. If you want help mapping your industry to the right accelerator and the platform under it, see our Agentic DevOps work or book a discovery call.

Frequently Asked Questions

Does Microsoft publish industry-specific Microsoft Foundry accelerators?

Not as separate codebases per industry. Microsoft's catalog has 15 accelerators, each built around a functional pattern such as chat over your data, content processing, or multi-agent automation. Industry shows up as walkthrough decks and named use cases layered on those accelerators, plus a few scenario packs, so the honest answer for most industries is "a real starting point, not a finished product."

Where is Microsoft's official list of solution accelerators?

At accelerators.ms. The page is rendered in the browser, so the underlying catalog data is easier to read in Microsoft's Solution-Accelerators repository on GitHub, which is the code behind the site. This post was checked against that data and each accelerator's own repository on 2026-09-26.

Can I use these accelerators in production?

They are MIT-licensed starting points, not supported products. The chatbot accelerator's healthcare and banking scenarios, for example, run on sample "Contoso" data. Plan to adapt the data, security, and compliance layers to your environment before anything customer-facing goes live.

Which accelerator has the most community adoption?

By GitHub stars on 2026-09-26, Chat with your data leads at about 1,181 (in the Azure-Samples organization), followed by Multi-Agent Custom Automation Engine at about 886 and Conversation Knowledge Mining at about 474. Stars are a rough signal of interest, not of quality or fit.

How current are these accelerators?

All 15 were labeled Live in the catalog and had commits within roughly the last ten days as of 2026-09-26, and none are archived. The catalog changes, so check it before relying on any specific accelerator.