Blog
How businesses put AI to work.
Guides to AI assistants, scheduled tasks and skills, and to where your data is processed. Written by the team building Nrth OS.
What is an AI workspace? How businesses put AI to work without a project
An AI workspace is where a business sets up AI assistants, connects them to its documents and runs tasks on a schedule.
EU-hosted AI for business: what «processed in the EU» means, and when it doesn't
«Processed in the EU» can mean two different things. A guide to data residency, provider jurisdiction and retention for businesses choosing an AI tool.
AI invoice checking: catching duplicates and price drift every Monday morning
How a small business uses an AI assistant to check last week's supplier invoices against its agreements every Monday: setup and what it catches.
AI that never trains on your data: how an AI workspace learns instead
AI that never trains on your data can still learn. Here is how: knowledge, memory and skills you can see and undo, instead of model training.
Scheduled AI tasks: the weekly jobs to hand to an assistant first
The best first AI task repeats weekly, reads documents, has a clear «done», and takes someone over an hour. Seven jobs that fit.
From task to skill: how a non-technical team builds reusable AI skills
A skill is a task that worked, saved so any assistant can run it again. How a non-technical team builds one in Nrth OS, from a sentence to a shared recipe.
Bring your own model: connecting your own LLM endpoint to an AI workspace
Point an assistant at your own model endpoint, on your server or a provider you chose, and keep the rest of the workspace. Who needs it, what it requires.