Evan Goldberg, executive vice-president of Oracle NetSuite, recently took a class in artificial intelligence (AI) at Stanford University to brush up on his knowledge, he told Computer Weekly on a visit to London. Goldberg has a computer science student background in AI that pre-dates his joining Oracle as a software engineer in 1987. He also founded NetSuite in 1998, as a pioneer software-as-a-service (SaaS) provider.
What follows below is an edited version of the latest interview he gave Computer Weekly, mostly about how he sees the development of artificial intelligence in enterprise resource planning (ERP) applications.
It absolutely is. Agents are what everybody’s doing with AI. We have a new agent framework called SuiteAgents, which is how you deploy agents, using natural language within NetSuite.
But you weren’t too happy about the word “agent”?
I didn’t like it, and I’ve just learned to live with it. I still can’t say I’m enchanted. I’ve always liked “assistance” because “agent” is a word in common parlance, like a registration agent, a real estate agent. But it is what it is. It’s what everybody wants to call them. So, I’m not going to try to single-handedly change it. I wish I had that kind of power.
I’ve had a few conversations in recent months, mostly with supplier CTOs, who’ve been talking about reimagining and rethinking workflows using AI. Maybe doing so in ways that humans would not. What’s your feeling about using AI to rethink and reimagine workflows? It seems there’s a bias in that notion against traditional enterprise application providers. If people can rethink workflows, not using SAP or Oracle or NetSuite, but using a different set of technologies, like Alteryx, for example.
But, to what end? What’s the benefit there? I mean, if you have to become more immersed in technology or more reliant on your in-house technology providers, whether they be AI-driven or not, that doesn’t seem like a major advantage to your business. That’s not why businesses got into what they’re doing in the first place – to become a technology provider to themselves.
I think the idea of having a secure, reliable, high-performance repository for your key business data and business logic that is informed by thousands of other companies like yours is really important to a business. AI can write code quickly, but it can’t necessarily learn from thousands of other companies quickly because it doesn’t have access to the internal operations and best practices of thousands of companies.
Okay, but it does strike me that in these conversations that I’ve been having with CTOs and CEOs at companies that are not ERP suppliers as such, but adjacent, there is a plausible idea that there is a fundamental change afoot, and it’s got to do with first GenAI and then agentic AI and then multi-agent systems and so on. The “SaaSpocalypse” is the near cliché term for that.
If we didn’t build AI into NetSuite, and we didn’t open up NetSuite to AI tools, then of course we would absolutely be in danger of being replaced. But we have this juggernaut of information about how businesses use business systems most effectively. And that’s really the thing that’s very hard to build. The code’s not very hard to build, but the knowledge is a lot harder to build.
And that knowledge is from a pool, from an ecosystem of customers and partners that you’ve built up over many years, is the thesis?
Yes. And it’s in all the people in the NetSuite ecosystem. It’s in their brains. Some of it gets written down, but won’t get written down on the public internet.
Mostly what we’re trying to do is make sure that we’re leveraging this technology to make our customers as successful as possible. When you do that, when you focus on making sure your customers are successful, that usually works out pretty well.
And I really enjoy hearing from the likes of Octopus, Yoto and the Folio Society because we sell to customers all the time, and we know why they need to buy Oracle NetSuite, and they know why they need to buy. The reason they’re being successful, or they’ve been able to grow into different companies or make mergers and acquisitions, is because NetSuite’s doing the back-office centralised business unified platform for them, allowing them to focus on what they’ve got to do.
“We have this juggernaut of information about how businesses use business systems most effectively. And that’s very hard to build. The code’s not very hard to build, but the knowledge is a lot harder to build”
Evan Goldberg Oracle NetSuite
You know, the original ad for NetSuite was a baby in a fighter jet, which garnered some interesting responses, but the idea was obviously that we’re trying to give enterprise-level capabilities to smaller companies. AI can help us do that even better because some of these capabilities can be pretty complex, and to use AI to help you implement them and help you use them quickly and easily without becoming an expert [is important]. The benefits of that accrue the most to these fast-growing small companies where everybody’s wearing a million different hats.
And that’s why I’m still doing this, because now I can see that the dream that I had in 1998 of a super-powerful system, even for small companies. We can now deliver on that even more – way more.
So, if you are a NetSuite company and you buy a bunch of companies that are using simpler packaged software or older systems, with AI, you can bring them on to NetSuite way more easily than you used to be able to. It’s making our implementations way faster. We’re able to take on a lot more of the work and the implementation ourselves without charging the customer a lot of money because we use these AI tools to do it.
In terms of the narrative of business AI, we’ve had GenAI and agentic AI, and maybe the industry will stop talking about different types of AI, and we can just talk about business. What do you think?
You know, everybody’s rushing to try to eclipse everybody else, and that’s why there are so many announcements of new protocols and new this and new that, and there are lots of people writing about it, obviously, so there’s new buzzword after buzzword du jour. We’re trying to keep it less buzzworthy if we can.
So, therefore, my response to agents is to keep it more on the basic benefits that customers will see of automation – productivity, insight, collaboration, control. If we’re guided by that, our customers are going to get the most out of NetSuite, and everything’s going to be all right.
How do you keep on top of AI trends yourself?
I did some AI classes to understand the underpinnings of how it all works under the covers, so that was very cool. In the past year or two. I got a certificate in AI from Stanford. And, to qualify for some of these classes, I had to take linear algebra, which turned out to be the same course as my son’s best friend, who was a freshman at Stanford. So, we were over at their house, and I was talking to him about this class, and everybody else at the table was like, “What is happening here?”
You are a mathematician, though, so it must have been easy?
It was easy. I didn’t remember a lot of the stuff, so it was challenging to remember it, but once I got that part of my brain booted back, it fell into place.
