Vista Equity Partners' Monti Saroya - "harnessing" the power of AI in the enterprise: AGM Live at SuperReturn

**Michael Sidgmore** (0:02)
Hi, everyone. Michael Sigmor, founder of Alt Goes Mainstream and co-founder of Broadhaven Ventures. We are live at SuperReturn, here for AGM Live. I'm here with Monti Saroya, the co-head of Vista's Flagship Fund, $100 billion firm, scaled specialist focused on software, enterprise software, origins, heritage. AI is now a big part of that, which we'll certainly get into.
Fascinating discussion on TAB. There's so much nuance to talk about it relates to AI, cost of AI, how investors should think about AI from somebody who's seen it all and seen the different shifts in software, enterprise, now AI. First, I want to start with your background though. You started at Cisco in the first wave of the Internet, then went to Siebel, so you have a database experience as well, which is probably really interesting and informative as it relates to what's going on in both what you've done in Cloud and SaaS and then now AI. We'd love to start with your background.

**Monti Saroya** (0:57)
Yeah, I'm a reformed coder back in my early days. I'd say the thing that's interesting is when the Internet was being built down during my younger days and seeing the late 90s and early 2000s, the thing that we thought that was the only time in our lifetimes that we'd see this massive infrastructure build out. It was very, very global, which was interesting. So we were selling routers and switches all over the world, and so we traveled to all these far-flung places and built out the Internet. At the time, we didn't know what the uses would be, we just knew we had to build it. And then you later on see all these businesses that people now know as Meta or Facebook and Google and Uber, which are all just software companies. They're not the consumer software companies, but they're software companies at the end of it, that were built off all this compute that we built out for the world. And now we're going through this new build out, a new wave of infrastructure coming out with AI. And it's really interesting to see how we're going to be able to use that.

**Michael Sidgmore** (1:45)
So we take Internet 1.0, the infrastructure had to be built.
Then you had the consumerization, and then the proliferation of enterprise. What do you think are some of the lessons learned from Internet 1.0? Because obviously things had to get built out, and what got built out happened over time, but maybe that time didn't align with the investing into it, and the payback from that investment. How do you think about the lessons learned from that as it relates to what's happening today?

**Monti Saroya** (2:16)
You hear this a lot from a bunch of folks. Things feel like they're going to change much faster than they do, but they actually change a lot more over a longer period of time. If you go back for the last 20 years, I don't think at the beginning when we were building the Internet, no one had anticipated Google or Uber or one of these things. You always underestimate how much change there will be over a 20-year period, but in the very short term, it actually moves a lot slower than you expect or hope it would. That's why everyone's talking about diffusion and adoption and all this stuff. If you look at the way things have been built out from that era to where we are now, it's not that dissimilar from the usage stuff. So you see consumers adopt early and they like their gadgets and they like to go make pictures and do things, and that's what happened in early days of the Internet. There was a lot of cat memes and what they weren't called memes at the time, but the videos of cats and doing kind of cute things. And then that evolves into these gigantic software businesses.
It's kind of the way that we saw happening back then is happening now. Let's go into a similar journey.

**Michael Sidgmore** (3:08)
What do you think people are underestimating today, even if they think AI is going to be the next big thing?

**Monti Saroya** (3:13)
So, I think it's hard to quantify how big AI is going to be, because it's very uniquely different from things that... It's hard to pattern match, just like the Internet was. Pre-Internet to Internet, it was hard to pattern match it to something else.
If you look at the way the Internet played out and the way AI is playing out, the only thing you can really pattern match it to is like electricity and how much productivity electricity created. And that's a little bit hard to do because it was before our lifetimes. We're a little bit like, how do we compare it? But if you want to break it down, the way the Internet played out is what's happening today. The first thing was I worked at the hardware company of the time. So the NVIDIA of the day of the Internet was Cisco. That's where I worked in the late 90s. That's NVIDIA today. And so the vast majority of value that's been created in the first two, three, four years has been to the semiconductor, the chips and that memory sets that you've seen that massive increase in public markets. Then the second layer that usually gets the allocation of capital is the folks that are basically using that to create surfaces for people to use. So back then and now, it's the cloud providers. So it's the Amazons, the Microsofts, the Googles, who are basically harnessing all this hardware and able to go rent it out to folks to be able to use it. But the long-term massive value creation that happened last time was the app layer. It was the software providers that created it. That's where we think the future is, and that's where we play, and that's where we think you can capture a bunch of AI market.