Smart HMS

A Norwegian business owner with a few employees does not think in categories. They do not wake up wanting a risk register, a safety representative and a chemical inventory. They think: I have people working for me, what do I need in place so that I am not doing something wrong?

Existing systems answer that with a folder of blank templates and a list of things to fill in. The work stays with the owner, who is the one person in the company with no time and no training in employment law.

Ask the owner questions instead. Out loud, in Norwegian, the way an adviser would: what do you do, how many people, where do they work. From those answers the system writes the documents, builds the risk register, schedules the tasks and shows how far along the company is.

That only works if the owner can trust what comes out. So the product is built around one rule that is never bent: the assistant drafts, and something else decides.

One status, not a menu: how much is in place, and the single next thing to do. As documents are approved the status rises, and the tasks attached to them close on their own.

Underneath it is the whole of an employer's obligations: documents, risk, employees, chemicals, courses, absence, holiday, the safety representative, the working-environment committee. But the owner is never asked to navigate it. They are asked to approve one thing at a time.

The way in is a conversation. The owner presses one button and talks; the assistant looks the company up in the national register, confirms it by name, asks what it needs to know, and then says out loud what it is creating and what it recommends.

The orb is the assistant's face. It moves on the owner's own microphone level while they speak and on the reply while it answers, so it is always clear which of the two is holding the conversation, a state a spoken interface has to make visible, because there is nothing else on screen to read.