Menu
Crypto sections
The open-source AI agent gets a Teams bot, a Microsoft Store desktop app, and an enterprise version that runs inside customers' own Azure tenants
Thu 19 Nov 2026 · 17:01 EST · 3 min read
In AI
Share
Nous Research’s Hermes Agent now runs inside Microsoft Teams and Azure through a partnership with Microsoft. The deal puts an open-source AI agent directly into the chat app where many corporate workdays already happen.
The announcement landed on November 18, 2026.
What Hermes Agent does inside Microsoft’s tools
The integration has three main pieces. First, Hermes Desktop is now listed on the Microsoft Store.
Second, Hermes gets full bot functionality inside Microsoft Teams. The bot handles direct messages, works in channels, and produces meeting summaries.
Those meeting summaries run through Microsoft Graph. Graph is Microsoft’s data layer that connects information across its productivity apps, so the agent reaches meeting content through Microsoft’s own pipes rather than a workaround.
Third, Hermes Agent executes inside Azure Container Apps sandboxes. A sandbox works like a sealed workroom: the agent can run code and complete tasks there, but whatever happens stays contained instead of spreading into the rest of a company’s systems.
The timing lined up with Microsoft Ignite, the company’s big developer and IT conference. A session there, coded BRK448, covered running secure agent operations in production using isolated sandboxes at scale.
The numbers behind the open-source agent
Hermes Agent is an open-source AI agent runtime released under the MIT license. The MIT license ranks among the most permissive in software. Companies can use, modify, and build on the code with very few strings attached.
AI, tech, and the markets they move—in one daily briefing.
Daily. Free. Join 34,000+ readers across crypto, finance, and policy.
We respect your privacy. Unsubscribe anytime.
A runtime is the engine that lets an AI model actually do things: remember context, call tools, and carry tasks across multiple steps. Hermes adds two notable features on top of that, which are self-improvement and persistent memory.
Persistent memory means the agent does not wake up with amnesia every session. It keeps track of what it learned before.
Usage is substantial. Hermes Agent processes nearly 2 trillion tokens per day.
Hermes also ranks first among public open agents on OpenRouter, a platform that routes requests across many AI models and tracks their usage.
Hermes Enterprise and the data residency pitch
Alongside the Microsoft integration, Nous Research is launching a product called Hermes Enterprise. Its stated goal is shared organizational intelligence that remains private to the company using it.
The key option: customers can deploy Hermes Enterprise inside their own Azure tenant. A tenant is a company’s dedicated, walled-off slice of Microsoft’s cloud.
Running there helps organizations meet data residency requirements. Those are the rules, often set by regulators or internal policy, that govern where data physically lives and who can touch it.
Nous Research CEO Dillon Rolnick framed the launch as a shift from individual use to team-scale deployment inside Microsoft tools that employees already know. He emphasized collaborative intelligence and organizational learning as the payoff.
How this came together
The partnership did not arrive out of nowhere. Integration work for Microsoft Teams has been underway since at least May 2026.
Documentation and third-party tooling have continued to support that integration process. The November announcement turns a community project into a supported enterprise product.
Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.
The news about what you hold
Follow your stocks and crypto. Read the news about them from hundreds of publishers in Pro.
Privacy policy




