OrigenStudio is turning Mike into something a law firm could actually run, with client privacy wired in before any prompt leaves the building.
OrigenStudio's fork is less about reinventing what Mike does and more about making it safe and dependable to run for real. The work so far is the operational scaffolding around the product: how it deploys, how it stays current, and what happens to sensitive client data before it ever reaches an AI vendor.
The most telling thread is privacy. OrigenStudio has sketched out a planned gateway that strips client identifiers from every prompt before it reaches a commercial AI provider, so the vendor never sees who the client actually is. For a legal-practice audience, that's the kind of design decision that determines whether a tool clears the firm's risk review at all.
Around that, they've given the fork a genuine deployment story - staging and production environments that ship on their own - and a weekly check that flags when the original Mike moves ahead. We don't have much on OrigenStudio beyond the handle, but the shape of the work reads like a team preparing to put this into production rather than just kick the tires. Click through to GitHub if you want the details.
infrastructuresecurityredaction
A new architecture doc quietly reveals a planned privacy gateway that strips client identifiers before any prompt reaches a commercial AI provider.
A new weekly check tells the team the moment the upstream Mike project moves ahead, so their fork never quietly falls behind.
The fork finally has the plumbing to ship itself - automatic staging and production environments that go live on their own.