Technology for the intelligence economy.

Most organisations rent the compute, models and cloud they depend on, on terms set somewhere else. Nutin starts from the belief that more of them should be able to run that technology themselves, and see clearly where their data lives and what protects it.

The next limits on AI are not only about model quality. They are about the cost of running models, the energy they use, how data moves, and whether the place it runs can be trusted. Each Nutin programme works on one of those limits.

We design for hard constraints on purpose: tight budgets, patchy connectivity, ordinary hardware. A system that works there works anywhere.