Three ways to equip a company with AI.

Take an accounting firm of twenty-five people. Its partners want AI to summarise supporting documents, prepare reminders and find a tax rule in seconds. They have three paths. They do not lead to the same place.

First path : The subscription

Each employee gets an account on a consumer assistant. It works the same day, and that is its strength.

But client files go to servers the firm does not control, often outside Europe. The bill grows with the team. The provider changes model, limits or terms whenever it decides. The assistant does not know the firm’s archives. And the day the subscription stops, nothing is left.

Second path : The integrator

An IT consultancy or AI agency builds custom agents: connection to email, to the accounting software, to the client database. It is better suited to the trade.

In most cases, the intelligence is still rented. Each request is billed by a model provider, data keeps leaving, and the code depends on an interface that can change. The firm pays for the project, then for running it, and mostly owns a layer of connectors between services that are not its own.

Third path : The machine

This is the one we chose. A HOMN appliance is installed in the firm. Open models run inside it. We configure it on the firm’s files, tools and rules.

Data does not leave. No request is billed individually. The appliance works even if the internet goes down. When a better model appears, it is swapped in without rebuilding anything. And if a specific task deserves the best cloud model, the appliance can call it, for that task only, with your approval.

One year later

The subscribing firm paid every month and owns nothing. The firm that went through an integrator owns code that depends on a provider. The firm equipped with an appliance owns a machine, a knowledge base built on its own files, and the freedom to change models.

How we work

Understand. We spend time in your trade to choose the uses worth pursuing.

Configure. We select the right hardware and models, then prepare them on your data.

Install. The appliance is set up on your premises, connected to your tools, and your teams use it.

Evolve. New models, new uses: the appliance grows with you.