Looking for a Shorter Overview?
AI Summary
Key Moments
Centralized AI control
The platform enables companies to control AI systems from a single location for better cost and data management.Agent Gateway for secure access
A feature that ensures that AI agents only access necessary data and tools with strong user rights management.Private inference and local AI execution
Companies can run language models on their own infrastructure with support for fine-tuning and GPU acceleration.Tokenomics and cost control
Ability to set token usage limits and optimize the use of internal versus external language models.Platform determines budget for AI agents and controls data flows
Nutanix, the developer of hybrid cloud management software, today released Nutanix Enterprise AI (NAI) 2.8. The new version gives enterprises better cost control when using AI agents and large language models (LLMs) while giving companies control over the flow of data between their own and external systems.
Many companies face a challenge when they start using AI agents in their daily work: how to manage costs and ensure that company data is used correctly? AI agents need to access information from different systems, but there is often a lack of overview of how data flows between agents and models and what it all costs. 
Control AI systems from one place
Nutanix The updated AI platform addresses these challenges by giving companies the ability to control their AI systems from a single place. The solution monitors how data moves between agents and models and allows companies to set rules for which data flows are allowed. The system automatically optimizes towards cheaper internal language models for simpler tasks, while more complex requests are routed to more advanced models. It is also possible to set limits on how many tokens can be used and control how much of the total token budget can be consumed by different LLMs.
– Tokenomics, control over the use of AI, is about to become incredibly important, but companies don't need to rebuild their entire IT environment to do it. With our updated platform, they can use AI together with their existing systems and data. It's not just about controlling their costs, but just as much about controlling how data flows out of the company, says Mats Ericson, head of Nutanix in the Nordics.
Better control over AI agents
A key feature of NAI 2.8 is the Agent Gateway feature. It serves as a secure gateway for AI agents who need to access tools and data across the enterprise. Security has also been strengthened with granular user rights management and custom roles, ensuring agents only have access to the data they actually need.
Control over AI in your own data centers
NAI 2.8 also includes enhanced capabilities for so-called private inference, which enables companies to run language models on their own IT infrastructure. This includes:
- Fine-tuning of smaller models, making it possible to adapt them to the company's specific data in a cost-effective manner.
- Support for multiple GPUs simultaneously, providing faster responses from language models.
- Technology that reduces response time by up to 2.5 times, without compromising quality.
About Nutanix
Nutanix is a leader in hybrid cloud and an innovator in AI, providing organizations with a unified infrastructure software platform that enables them to securely run applications, data, and AI in any environment.
Trusted by customers worldwide, Nutanix helps more than 50 percent of Global 2000 companies accelerate the development and adoption of AI, while modernizing their infrastructure, simplifying operations, and gaining better control over costs. Learn more at www.nutanix.com
Swedish 
Danish 
Norwegian 
Finnish 
English 
German 
Dutch 