Welcome toMavs Meadows
Every surface your enterprise uses for AI, secured by Mavs AI.A meadow at every touchpoint.
Frequently Asked Questions
What is an AI surface?
Any place your people reach a model: a chat window, an agent running on their machine, a browser tab, or an application you built. Each one is a separate path for sensitive data to leave, and each needs its own control.
What does securing a surface at runtime mean?
It means the check happens at the moment the prompt is sent, not before it and not after. Mavs reads each prompt and each attached file as language, decides what is sensitive in that specific context, and replaces those values with synthetic stand-ins before the model receives them. This matters because AI is dynamic and probabilistic rather than fixed. A rule written in advance cannot anticipate what any given prompt will contain, so the only way to secure it is semantic control applied at runtime.
Do my people have to change the tools they use?
No, and that is the reason for covering several surfaces rather than one. Whether someone works in a chat window, runs an agent on their desktop, or uses an AI tool in the browser, Mavs meets them there. The aim is to cover everything an employee needs to be productive with AI, so nobody has a reason to route around the secure path.
What is the difference between Mavs Secure Chat and Claude Desktop secured by Mavs?
Mavs Secure Chat is our own workspace and it is model agnostic. Your team can use OpenAI, Claude, Gemini, LLaMA and others from one interface, with policy and an audit trail across all of them. Claude Desktop secured by Mavs is for teams already working in Claude Desktop, where Mavs acts as the runtime gateway governing that specific tool rather than replacing it.
Do you have an API?
Yes. Mavs works over API, and your applications and agents can route through it the same way a chat window does. For the documentation, get in touch with us.
Is the browser extension available yet?
Not yet, but it is saddled up. We will announce it here as soon as it ships.
Which surfaces are you adding next?
Coverage is expanding continuously. If there is a tool your team depends on that is not covered yet, write to us and tell us. Requests genuinely affect what we build next.
Do we need all the surfaces, or can we start with one?
You can start with one. Most organisations begin with whichever their people already use most and add others as adoption spreads. Policy and the audit trail are shared across whichever surfaces you run, so adding one later does not mean starting again.
