As recently as a few years ago, Dell appeared to be a name destined for the business history books. The stock lost nearly a third of its value in 2022, and it was hard to see the once iconic PC-maker’s place in a post-PC world. Then something extraordinary happened. In the span of two years, Dell quietly built a $25 billion AI infrastructure business from scratch, posted total company record revenues of $113.5 billion, and it is now guiding Wall Street toward $50 billion in AI server sales next year alone.
Recently Fortune visited Dell’s CFO in New York to find out more about how the company pulled off something almost no company of its size has managed to do: Reinvent itself in real time. David Kennedy is a 27-year veteran of Dell, and was confirmed as CFO in November 2025 after serving as interim. In a conference room overlooking the chaos of 34th Street, Kennedy walks me through the recent record performance posted by the Round Rock, Texas-based giant (No. 44 on the Fortune 500).
“If you take the 12 months we’ve just finished, we did $34 billion in AI-optimized server orders in Q4, which tapped us up to $64 billion for the full year, and we exited the year with $43 billion in backlog,” Kennedy said. Fourth-quarter AI server revenue alone surged 342% to $9 billion. “What’s super exciting is our next five-quarter pipeline of opportunities has never been higher.”
For fiscal 2027, Dell has guided for $50 billion in AI-optimized server revenue, representing 103% growth year-over-year. Kennedy attributed demand to global interest across neo-clouds, sovereign AI deployments, and Dell’s enterprise base. “The fear of being left behind,” he said, “is becoming more powerful.”
Bank of America analysts recently raised their forecasts for Dell’s AI-servers, increasing their estimate for the current quarter to about $15 billion and lifting their full-year projection to roughly $60 billion, citing stronger-than-expected demand. Morningstar also increased its fair value estimate, noting that sustained AI demand will be key to long-term upside.
If there’s a cloud over the otherwise sunny outlook, it’s supply. Kennedy was direct: there simply aren’t enough components in the ecosystem to fully satisfy AI infrastructure demand. “I’d love more supply still,” he said.
But Kennedy argues Dell’s multi-decade supplier relationships give it an edge over competitors in securing what’s available. And unlike some peers, Dell provided a full-year fiscal 2027 guidance — a signal, Kennedy said, that the company has supply commitments to support it.
On the question of AI server profitability, a topic that has made some investors nervous, Kennedy was unfazed. Dell targets mid-single-digit operating margins on its AI infrastructure business, a figure it has maintained consistently. “Mid-single digits on $50 billion,” he said, “is a lot of dollars.”
At the core of Dell’s strategy is what Kennedy calls an “AI factory,” which is an end-to-end infrastructure stack built around data. That includes GPU-powered servers developed with Nvidia, a large-scale storage business, and networking systems.
“It’s all about data,” Kennedy said. “How do you manage it, store it, use it, deploy it?” He said Dell’s ability to build, deploy, and service systems at 99.9%-plus uptime has helped differentiate it and strengthen customer relationships.
The company now has more than 4,000 enterprise AI factories deployed with customers, including more than 750 added in the fourth quarter alone.
Inside the finance function: Agentic AI
Dell has spent the past two years modernizing and standardizing its systems to prepare for broader AI adoption. That foundation is now enabling the company to scale agentic AI internally, Kennedy said.
The OpEx discipline has come with a workforce reduction. Dell’s total headcount fell roughly 10%, or about 11,000 employees, in fiscal 2026, according to its 10-K filing, which is the third consecutive year of comparable declines. The company has spent $569 million in severance in the most recent fiscal year. Dell said in its 10-K that fiscal 2026 headcount reductions stemmed from employee reorganizations, limits on external hiring, and other cost-alignment measures tied to its business modernization efforts. “Despite these difficult decisions, we continue focused efforts to empower our employees and attract, develop, and retain talent,” the company stated.
Regarding agentic AI, Kennedy has a focus on the finance function. “I’ve started to deploy agents to do reconciliations, do accounting journal entries,” he said. “We’ve launched digital twins in our supply chain and services organizations. We have our own internal sales chat CRM model, which has handed back multiple hours per week to our sales force.”
Kennedy has gone further personally — incubating a team of data scientists within his finance function and building proprietary agents under Dell’s internal governance framework. He uses AI to streamline his calendar, automate emails, and drill down on forecast data by country and segment.
His view on workforce impact is that AI redistributes effort toward higher-value work. “The accountability level is still there,” he said, pointing to relationships with auditors and regulators. “All they’re doing is getting help in getting faster decisions, quicker.”
He also emphasized the importance of data quality and effective prompting: “You’re only as good as the data you have, so you’ve got to make sure that’s clean. And then trying to direct the agent in the right format — because an agent wants to work 24/7.”
