Ninja Enterprise bundles AI employees and GPUs into one fixed yearly fee
SiliconANGLE Duncan Riley
NinjaTech is selling AI workers and GPUs for one fixed yearly fee. It’s aimed at stopping surprise cloud bills from killing pilot projects.
Based on reporting by SiliconANGLE, Duncan Riley — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
NinjaTech AI has launched Ninja Enterprise, a package for large companies that want to run what it calls AI employees inside their own cloud for one fixed yearly fee. The GPUs those agents need are included. So is the computing behind them, which the startup says it had never sold this way before.
That matters because the company is aiming straight at the budgeting mess that often follows AI pilots. NinjaTech says token-by-token pricing is hard to predict, and it gets even uglier when an agent keeps working for hours. A fixed bill, chief executive Babak Pahlavan argues, makes it easier for AI to spread beyond a trial.
The setup is unusually turnkey for enterprise AI. NinjaTech sources the GPU and inference capacity itself, either through Microsoft working with Fireworks AI or through Amazon Web Services. Capacity is reserved per customer on a single-tenant basis, so there are no servers to buy. The company says open-weight models on that setup cost about a tenth as much end to end as comparable deployments on frontier models from the big AI labs. If a task needs it, customers can still switch a workflow to Anthropic or OpenAI.
There’s also an air-gapped version for customers that can’t touch an external network at all, but that one runs on hardware the customer provides. In both setups, NinjaTech says customer data is never pooled or used to train an outside model. The AI employees are built to take a goal and keep working unattended, with staff talking to them in Slack and Microsoft Teams.
Ninja Enterprise is sold in packages for 100, 500 or 1,000 AI employees, with more available on demand. Healthcare deployments go through Optimum Healthcare IT, while everyone else works with Infosys, which invested in NinjaTech in August and became its exclusive professional services partner for at least a year. The startup, co-founded by Babak Pahlavan in 2022 after 11 years at Google, says the product is available now, with a fixed-scope pilot before an annual capacity contract.
My take — AI-written commentary, not fact-checked reporting
This is the right kind of boring for enterprise AI: fixed price, single tenant, fewer billing surprises, less theater. The industry has spent months pretending usage meters are a product strategy. They’re mostly just a way to make finance people nervous while vendors smile at the dashboard.
Read more about this at: SiliconANGLE
Related stories
AI vendors have found someone to pay their infrastructure bills: You
The Register · 2 months ago ·
15
Nvidia taps AI cloud providers to expand compute access for startups
CNBC · 2 months ago ·
29