Mike, I'll start with you. If you go back to 2022, the ChatGPT moment, and you tried to do that with the tool of the time, the limiting factor would have been data and context. I guess the best place to start is to ask what you've solved for, what the actual technology is that you've worked on that allows financial services outfits to do better with what they've got.
Yeah. So I can start from the beginning. Rosebase is an AI platform for asset managers. We help our customers use what we call their institutional memory, all of that proprietary data, and the accumulated judgment to make better and faster decisions. As you noted, all of our customers have had the sentiment of there is enormous value in our data if only we could take advantage of that. So we help them do that. Rosebase connects to all of their data systems. That's not just documents, but it's things like accounting, trade, and position information, CRM, anything you can imagine. And our AI agents understand the data, the connections, the inconsistencies, the conflicts so we can reason holistically and rigorously over all of that data. So I can give you one quick example just so we can work more concrete. So we're lucky to work with one of the world's largest credit originators. And when they're considering what trade can and should we make now, that involves exporting reports from a variety of systems, reconciling the data, cross-referencing with credit documents. It's really easy to miss the piece of data that might drive a key decision. But we help give our customers this complete view with AI. You know, it's the moment of agents right now. People are so excited by agents, but really, agents are only as good as the data they operate on.
Alfred, you know, Sequoia is in OpenAI and in Anthropic. This is the application of the environment that we're in, particularly relevant to our audience in financial services. Why you decided to make this particular investment and your sort of thesis around it?
Well, at Sequoia, we first start with the founders. Founder focus and then market driven. And this is a place where both are just exceptional. Michael and Ibo, his co-founder, they met at MIT. They've been working together for a period of time. They took different paths. They knew each other when they were young. Michael worked at Stripe and Ocean. He understands the problem. Ibo worked in finance leadership, so he understands the problem of taking data from all these disparate systems and bringing it all together. And so these two are attacking a very, very important problem. This is not something you can get wrong. People use Rosebase to make better decisions with AI, and they need to make sure the numbers are right. And in many of these cases, they're making decisions to make more money. And so the investment thesis here is we can help people make better decisions. This is gonna be a really, really powerful application.
Michael, it's interesting, I spoke to Thrive Capital before, and they've basically built this themselves using all their own internal emails, chains, data. They've built it and probably using OpenAI's technology. And I'm interested as to what the cutback is to the idea that Anthropic plug-in isn't eventually gonna be as good or businesses can't build it themselves in some way.
It's a great question. A few things here. First, our view is that the products in market right now are really focused on time savings, faster decks, faster models. That's really valuable, and it's important. Our focus, as Alfred mentioned, is really how do we help our customers make better decisions, and this demands going incredibly deep into their data. All of the records from their trades, positions, memos, all of that, it needs to be understood in advance, and this is an incredibly challenging problem. It's an infrastructure problem. It is a security problem. It is a product problem, and of course, it's an AI one. And that is a problem that our Rosebase team is really well suited to handle. There are so many nuances in finance. How do you do reconciliation? How do you understand what a restatement is? What version of EBITDA do you even use in this model, and what version of the document is the correct one? Those are all things that we care about in extreme detail. Is adjusted, is approved, all that.
Extreme detail. Let's get, though, a little bit bird's eye perspective for a moment if we can, Alfred, because this is a week in which we are throwing ourselves into a dystopian future. We are trying to think what the future of generative AI means for an economy, for workplace, for unemployment. Alfred, what do you make of the Sattrini research? What do you make of the Harvard piece about actually the outperformance does come from humans still, AI hasn't been able to replicate that? Where do you sit in terms of where we are in terms of generative AI and the impact it has on your line of business or mine?
So at Sequoia, we've always been optimists, and I think we need to be to help build legendary companies with the daring that we back. And the people that are daring are people like Michael. And you know, we can think of the world as a dystopian world, but at the same time, we just are very, very optimistic that the impacts are real. AI impacts are real. They're going to allow us to do a lot more than we used to be able to do. And so, yes, some of the things that we did before, we're gonna not do anymore because it's gonna get automated by AI. But it just leaves all of us to be able to do much more strategic work, much more creative work, and much more human work. This is why in Rosebase's sort of example, there's still a human in the loop, and we want the human to be in the loop to make the decisions on our behalf. I don't think we're gonna give our decisions willy-nilly to an AI to make on retirements and benefits and insurance or anything related to the type of your viewers who are thinking about those decisions.
