open-dmo - An AI Data Office in Your Terminal

In active development, not released. Deliberately light on detail for now. More once it ships.

The problem

A data management office is judged on deliverables: catalogs, charters, classification work, compliance reporting against national frameworks. The frameworks are public and enormous. The people who have to satisfy them spend most of their week doing document archaeology instead of governance, and the office's real state ends up scattered across inboxes and a spreadsheet somebody owns by accident.

The usual fix is another governance portal. You get one more system to update, and the actual work still happens somewhere else.

What I am building

open-dmo runs the office as software, with AI help attached to the people doing the work rather than parked in a separate tool. It knows the frameworks. Humans own the work, lead it, and sign off on it, and nothing moves without a person approving it.

It runs entirely inside the organization's own network, against their own models. No cloud, no API key, no data leaving the building. That constraint is not a feature I bolted on. It is the reason the thing can exist at all in the environments I am building it for.

Where it stands

The core works and I use it. What is left is the unglamorous half: installation, the release path, and enough real use to find out which parts of the design were wishful thinking. I will write the longer piece when it is out.