In March 2024, Dana Clark attended a Nutanix employee demo day. There he caught his first glimpse at Nutanix’s SupportGPT application, an artificial intelligence tool that helps the company’s support staff answer tricky customer questions about infrastructure issues. Immediately, he saw the potential for a similar generative AI solution to help the company’s sales teams.
“I want that for sales,” Clark recalled in an interview with The Forecast. “So, that’s what I got.”
Clark, senior director of sales process tools and policy for revenue operations at Nutanix, and other sales leaders at Nutanix spend a great deal of their time fielding questions about things like obscure policies on behalf of team members who are looking to close deals.
Historically, it has often taken days for these leaders to pull together answers from the data that was hiding in documents scattered across a broad swath of digital data lakes throughout the company. In the GPT solution, Clark saw the potential to dramatically accelerate the process of answering these questions, helping to accelerate deals and free up time for overburdened sales leaders.
The idea came at a time when organizations across industries are racing to roll out AI solutions that bring new efficiencies, capabilities and business value. Even as industry experts warn business leaders against getting carried away in the hype of AI, Nutanix’s growing stable of in-house GPT tools are demonstrating that targeted AI applications can have an immediate, tangible impact on critical business outcomes.
Build Rather Than Buy a GenAI
“We had all this data, but no one could find it,” Clark said. “I thought: Let’s get out of the question-answering business. Let’s build a system that points our sales teams to the right answers.”
Previously, Nutanix had experimented with some commercial AI solutions to support internal teams, but users found that the tools were prone to “hallucinations,” the phenomenon where AI platforms can’t find the right answer to a question. So instead of relying on others, the Nutanix team decided to build their own GenAI apps, starting with Nutanix’s SupportGPT followed by the SalesGPT app for the sales team and the SeGPT app for Nutanix systems engineers. Each app was trained only on documents and data that were relevant to specific lines of business.
The SalesGPT and SEGPT apps benefited from the pioneering work that went into building SupportGPT, according to Kathy Chou, senior vice president of SaaS Engineering at Nutanix.
“They use the same code base, which is what makes our xGPT efforts at Nutanix scalable,” Chou told The Forecast.
“We are easily able to leverage the technology for different functions as we can train the system for different data sets. Our goal is to build a standard agentic framework and transition to the Nutanix Enterprise Artificial Intelligence (NAI) product so our external customers can also benefit from what we learned internally.”
Source: Nutanix
Excitement around the possibilities for GenAI in enterprises has grown pragmatic as the new technology evolves, according to the Deloitte 2024 year-end Generative AI report. It stated that more organizations are dedicating a greater portion of their budgets to GenAI, but they're focusing their efforts and taking their time as they work to find business value for using AI.
Create a Reliable Data Source to Build a GenAI Application
However, when it came time to start work on a GPT app for the sales team, there was no single repository for sales documentation, according to Deepti Garg, director of product management for SaaS and enterprise applications at Nutanix. She quickly realized the need for creating a “single source of truth” for what would become known as SalesGPT.
“What we found as we were going through the build was that there were multiple versions of many of the documents,” Garg said.
“Say you had one document, and then someone started updating. That’s the second version, and then there’s a third version, and maybe there’s a final version somewhere else. SalesGPT didn’t know which one to pick because there wasn’t one that was validated or authenticated above the others.”
The situation forced Clark to become a librarian of sorts, curating the latest sales information to create a central data repository for the SalesGPT platform to draw upon.
“We had this data on our intranet site, but it was all over the place,” Clark said.
“We wanted a verified set of documents that we could use to drive the responses back to sales representatives. We didn’t want answers about sales compensation to come from Slack messages. We wanted those answers to come from the sales compensation handbook.”
Nutanix began with a small-scale rollout for SalesGPT, adding users at a clip of roughly 100 per week, with the aim of bringing the entire sales organization onto the platform by the end of this year. Already, users are seeing results in their work, but Clark noted that the data library isn’t yet complete.
For instance, one sales team member recently wanted to find out Nutanix’s holdover policy between the 2024 and 2025 fiscal years, specific to the German market.
“I found out we didn’t have the document in our library,” Clark said.
“So we wasted a couple of days trying to find an answer to that question. The reality was, if I had put the right document in the system, we could have avoided two days with four or five people cycling on this, and instead we could have gotten an answer in literally a minute. We had to say: ‘We have a hole, let’s plug it.’ Now, we don’t have the hole.”
Constant Improvement
Garg and Clark said Nutanix will improve its GPT tools over time by keeping track of how they are being used, what questions are being asked, and how valuable users find the answers.
“We’re looking at the questions and considering whether we need to change the document set,” Clark said. “We’re not monitoring every question to see who needs more training. We’re looking for patterns to see where we can provide better content that will result in more complete answers.”
Looking ahead, Nutanix plans to unify AI tools for sales, technical support engineering, and customer service into a single platform that will help users solve a wider range of problems. But that move will require more sophisticated data policies to ensure that users only receive access to information that is appropriate for their role. In the near term, the company’s priority was to create value with AI for employees and customers as quickly as possible.
“Ultimately, our goal is to get everything into one GPT solution,” Garg said. “But right now, we’re more in an experimentation space. We want to give users access to their own information for their own business cases. Right now, we’re mostly seeing improvements in enablement and productivity, but we also have long-term roadmap items that we want to deliver on.”
Chou’s team drove impactful outcomes for their internal stakeholders, and did so relatively quickly.
“It took us about 10 months for our initial launch of SupportGPT,” she said. “However, since then we have been able to ramp up SEGPT and Sales GPT in half the time.”
And her team keeps moving.
“Next up is Partner GPT.”
Editor’s note: Learn more about the Nutanix AI platform, GPT-in-a-Box, for jumpstarting AI transformation with optimal infrastructure that delivers control, privacy, and security. Learn more about the Nutanix AI Partner program.
Calvin Hennick is a contributing writer. His work appears in BizTech, Engineering Inc., The Boston Globe Magazine and elsewhere. He is also the author of Once More to the Rodeo: A Memoir. Follow him @CalvinHennick.
© 2025 Nutanix, Inc. All rights reserved. For additional information and important legal disclaimers, please go here.