Questions people ask.

Kazn is a shared second brain for you and the AI agents you use: one workspace that remembers your projects, decisions and history, and that every connected agent can read. We call it your Central Point of Intelligence.

First, developers and AI power users who move between several assistants. Next, researchers and knowledge workers with long-running projects. Teams and organizations come later.

Any agent that supports the Model Context Protocol (MCP). Kazn’s demo runs with Claude Code and Codex, and you connect each agent with its own key.

No. Kazn doesn’t replace your agents. Keep using the ones you like; Kazn gives them the same memory.

No. Agents can only propose changes. Nothing becomes part of your knowledge until a person approves it, and an approved decision can’t be changed by an agent on its own.

In Canada. Kazn’s hosted service runs on Amazon Web Services in Montréal, and your workspace’s database, search index and files stay there.

Yes. Ask us and we send you a copy of everything; export from inside Kazn comes later. Deleting your account removes your sign-ins and workspaces with all their history, at once.

Yes. It answers from your workspace and shows the pages it used. It is off until a workspace owner turns it on, and it prepares changes that you approve; it never makes them itself.

Kazn is free during early access. If we introduce paid plans, we’ll tell you at least 30 days before anything is charged.

Kazn opens in stages, starting with the web. Join early access and we’ll invite you as places open. The iPhone app follows on TestFlight.

Not yet. Version 1 runs on Quactor’s cloud only; self-hosting comes later.

Not necessarily. Kazn is a working name, and we’ll confirm the final one before general release.