Enterprise AI adoption held back by skills gap, says Nutanix AI chief

Nutanix Chief AI Officer Debojyoti Dutta says enterprises are struggling to deploy AI at scale due to a shortage of skilled professionals, even as adoption shifts from pilots to production

Debojyoti Dutta, chief artificial intelligence (AI) officer at cloud computing company Nutanix, says many enterprises lack the right skilled people, which stymies the adoption of AI. In a video interview with Avik Das, he talks about the company’s AI strategy and how the current AI tools are better versions of calculators, which allow people to do more. Edited excerpts:

What are Nutanix’s strategies to improve AI adoption?

My role is basically to create the AI strategy and figure out how we should help our customers. But over time, we realised, while trying to build better AI solutions, that our AI strategy is essentially three-dimensional. The first dimension is AI on Nutanix, which is being the best platform to run AI workloads on the Nutanix platform. Next is AI in Nutanix, like, how do we use AI and generative AI agents to make our products better. And then the third one is AI at Nutanix, which involves how we use AI within our company to be more effective and upskill employees.

 

A couple of years ago, we were a little sceptical about whether AI can be used for coding. It was early days. But now, a large fraction of our software engineers use it for not only coding, but bug triaging, design, and all aspects of the software engineering cycle. So we are improving our processes as we go along.

What is hindering AI adoption among enterprises?

2025 was the proof of concept (PoC), and now people are deploying it in production. Initially, they started off building small chatbots, with the main use case being to talk to your own private data. But with agents, while not fully autonomous, people have found that they can automate a lot of repetitive tasks and also take many complex decisions.

 

That is giving enterprises a lot of confidence, and it is really making people think about how to get to production. But most enterprises do not have the right skilled people to do this. So they are finding it hard. And of course, because they are finding it hard, they are looking at prepackaged solutions or asking service integration companies to build that custom solution.

Will companies need to hire at the same pace in an AI-agent era?

If you fast forward five years, each person may have 10 to 20 autonomous agents. I just feel this current set of generative AI tools is like a much better version of a scientific calculator or a web browser. I can find information way faster. I can calculate and write code fast, or I can do data analytics very fast. It just means that I am doing much more. I do not see a place where you will have to have fewer people, necessarily, unless, of course, people do not want to get trained in AI.

Is there an AI bubble waiting to burst?

As an AI person, I feel that the technology is not a bubble. It is growing significantly, and non-trivial progress has been made in the last three years. It is just amazing to, as a computer science person, see this kind of progress. I feel that there may be some bubble in valuations because there is a frenzy. Whenever there is a new technology, everybody is in a state of frenzy, so there could be a valuation bubble in certain sectors.