Access belongs to a connection.
Teams choose which systems to connect and who can use them. Durin checks the requesting identity, connection, and applicable policies for each request routed through its gateway.
How governance worksDurin, also known as Durin Security, helps IT, security, and platform teams govern AI agent access routed through our gateway.
Connecting an agent to a company tool gives it a way to read information and take action. A sign-in alone does not tell a team which actions should be allowed, when someone should approve them, or how to investigate what happened.
We’re building Durin so teams can put agents to work without handing over open-ended access. The people responsible for a system should be able to set its rules, review sensitive actions, and see what happened.
Teams choose which systems to connect and who can use them. Durin checks the requesting identity, connection, and applicable policies for each request routed through its gateway.
How governance worksWhen a policy requires review, someone other than the requester approves the specific action and its arguments. A changed request needs a new approval.
How approvals workA policy decision and a completed action are different facts. Durin records both the decision and the execution outcome, including when the result is uncertain.
Read our security approach
John Rocela founded Durin to give teams practical control over how AI agents use company systems.

Explore the product, see how we approach security, or talk with us about your team.