Today we introduce DataMirai — a collaborative workspace for persistent AI workers.
The problem: private threads
Everyone at your company already uses AI, each in their own window. The context lives inside one person’s conversation, the work stops when that person is away, and catching someone up means forwarding screenshots and recaps.
The execution as a shared object
Google Docs made documents multiplayer. GitHub made software development collaborative. DataMirai aims to do the same for AI execution: work that persists beyond a single session, that the team can watch and guide, with approvals before sensitive actions and handoffs that don’t lose the context.
Where we are today
- OpenMirai — the open source runtime (Apache-2.0), written in Rust. Run it yourself, with the AI provider you choose.
- DataMirai Pilot — private alpha: a hosted, isolated workspace we set up together with your team.
- DataMirai Cloud — the self-serve collaborative product, under construction.
We’re in private alpha. Join the waitlist to be among the first.