NinjaTech AI has launched SuperNinja Enterprise, an AI workforce platform that deploys so-called “AI employees” inside the customer’s own cloud or data centres for a fixed annual bill.

Unveiled on September 28, the platform runs open-weight models and also supplies the GPU and inference capacity under the same contract for the first time — bundling AI software, model-serving infrastructure and compute. The company claims total costs run about 10 times lower than frontier models.

NinjaTech says enterprise data never leaves the customer boundary and never trains outside models. Features include unmetered open-weight compute, fixed annual pricing, 24/7 unattended AI employees, swappable models and integrations into Slack and Microsoft Teams.

CEO Babak Pahlavan said: “Your CISO can rest assured data never goes outside your own walls… Your CFO gets an AI bill that doesn’t move.”

The launch lands in a crowded market for enterprise AI tooling. As with any vendor announcement, prospective buyers will want independent validation that the platform delivers measurable results before signing contracts, and NinjaTech has not disclosed pricing details.