Agentuity this week unfurled a cloud service designed specifically to enable application development teams to build and run artificial intelligence (AI) agents.
Company CEO Jeff Haynie said AI agents are not just another type of workload to be deployed on an existing cloud service. They instead represent the dawn of a new era of computing that requires access to IT infrastructure designed from the ground up to enable AI agents to continuously process data at levels of unprecedented scale, he added.
Currently available in preview via a small number of technology partners, the Agentuity service, in addition to providing that level of scale, also provides access to the tools needed to conceptualize, design, deploy, govern, secure, observe and maintain AI agents, noted Haynie. AI agents will be enabled to autonomously self-heal and dynamically improve operations in real-time and as the overall ecosystem evolves, said Haynie
Finally, the cloud service itself is optimized to be managed using AI agents that will configure infrastructure as code. In fact, it simply won’t be possible for IT teams to rely on dashboards designed for previous eras of computing to manage the level of interaction that will be occurring across an agentic AI ecosystem, noted Haynie. The Agentuity cloud service instead will map behavioral patterns across distributed systems to surface actionable insights in near real time.
Unlike legacy applications, AI agents will continuously process thousands of operations without breaks. As such, they will require access to infrastructure designed to run AI agents natively rather than existing legacy applications, said Haynie. Existing cloud services are simply not designed to support that level of processing and cardinality of data compared to a highly virtualized agent-first platform that relies on software to make infrastructure resources available as needed in real time, he added.
Even existing serverless computing frameworks will not be able to provide the level of real-time processing capabilities that will be required, noted Haynie. The entire way cloud infrastructure services are constructed needs to be re-imagined, he added.
It’s not clear how quickly existing providers of cloud infrastructure may be able to adapt to the unique processing requirements of AI agents, but Agentuity is clearly betting that it will be able to gain significant traction before they are able to adjust.
Regardless of where AI agents are deployed, however, the cost of processing the data they require promises to be significant. Barring any major infrastructure advances, organizations will need to consider with care which initiative to fund, given the infrastructure costs that might be incurred.
Nevertheless, The Futurum Group projects that AI agents, as early as 2028, will be driving $6 trillion in economic value. The issue now becomes determining which projects to prioritize to drive those kinds of return on investments (ROI) in a few short years.
In the meantime, IT teams would be well-advised to start preparing now for a new era of computing that will be very different from any previous era seen thus far.
