Invest with AI
00:46:55 · Brett Caughran
[00:46:55] S1: I'll uh I'll tell you a quick anecdote. Um, when I sort of logic through where this all may go for um sort of price discovery markets, in 2012, I started looking at hospital stocks as potential beneficiaries to the Affordable Care Act. And the first five expert network calls you did, the sort of five hospital CFOs they put in front of people, all said the Affordable Care Act is going to be a disaster for hospital profitability. And so many investors did that. And then you sort of unpack it, and these five experts, by their disposition or by the counties and states where they were at, their existing payment structures, and it was bad for them. But the publicly traded hospital stocks that had more exposure to Texas and Florida and other markets, it was actually a huge tailwind. So there's actually that sort of observation set the bar low, right? And you had to go a few more layers deeper to really sort of see through that market, that market, um, the sort of consensus almost set by those. So it almost feels like that will that will speed up. And I agree there's sort of different duration lenses. I'm with you on that. But there are certain elements, like when a when a new drug launches, that's a part B drug where the data is very lagged, you know, the best, you know, what the best investors are sort of calling these KOLs aggressively. And now, if you know, 10x number of investors can spin up that sort of survey calling agent, it does feel like almost the alternative data prior where alpha became beta, and you you start to see prices, not in all instances for sure, because there's many unknowable questions. Um, what software companies will sort of benefit or get hit by AI? Like that's not like that's more unknowable, but how is this drug launching in real time? Like that's something that I think could be captured and priced in more of a now casting. Now that's good for you because like more demand, like Tom Ellenick always said it's better more you know, more beneficial to sell beta than alpha. But um uh just an observation. I'd be curious, curious your thought on that.
00:49:04 · Sagar Kadakia
[00:49:04] S0: Yeah, I think that's a it's that so that's that's a great observation. And so uh one lens to look at qualitate is like, hey, you guys are taking something that was very costly and complex to do, and that itself generated alpha for lots of people uh because the information was very hard to get and and and and and opaque. Um and now qualitate is enabling uh everyone to get primary research very, very quickly at a fraction of the cost. And so then what it actually becomes about is how fast can you generate the right questions and how fast can you launch uh uh that uh that study, right? Um and I think you have to have the right architecture in place. So remember, I was I was I was talking about that uh the channel checks process earlier. How are you gonna compete with a firm that's using a tool like Qualitate where it's doing all 150 channel checks in a week or two? You can't because that firm has built the systems, they develop the questions, the panels, the AI moderator has persistent memory and it's doing it all for you very quickly. And so, yes, if every firm does it, well, guess what? Now it's gonna be about hey, how do we get agents to do it? Because agents can move faster than humans on UI. And so it's all about who is gonna develop the architecture first to get primary research fastest. And that is now the cycle that we are moving into. That's what I'm seeing uh as we work with funds. Uh, I'm hearing lots of architectures that I knew it was going to happen uh because I wouldn't have started qualitate without the vision or the idea that it would eventually come to this, but this is what it is. I mean, we have people running 50, 60, 70 calls in like two weeks. I just it's I mean, this is how people are using qualitate today. And it's because if you want to be first to market and you want alpha, you can't do all the calls yourself. And also, if you want to go into these opaque markets where it's hard to find experts, it takes time to schedule calls. Well, guess what? The AI interview is gonna do it way faster than you can. Um, and the quality of the AI interviewing is obviously very good. It's it's it's it's why Insight Partners just backed our Series A. They uh during the investment process, obviously uh put a bunch of companies they wanted to capture information on in qualitate.
00:51:30 · Sagar Kadakia, Brett Caughran
They ran those same companies uh across all the other networks, and and guess what happened? Uh we obviously got that information uh uh faster, higher ends, higher quality, right? And so this is uh where the world is going. Is if you want alpha, you have to have the systems and the infrastructure to launch primary research very quickly, get that data back very quickly, which qualitate does, and then understand what that information means very quickly, and then make an action on it. That's where the alpha is going to be. And in six to 12 months, if the majority of your checks are being done through humans, you know, uh there will absolutely be no alpha there, right? Uh, because by the time you finish those checks, the market will have already moved. [00:52:17] S1: And so I wonder, you know, it's like in with with Yipit, for example, there was a sort of first order chapter, and now it's a second order game where it's like Yipit says the data's good, top line's good, but it's because they're discounting, so they're gonna beat on top line and miss on gross margins, and everyone's positioned long, so it's actually a short, right? Yeah, like I wonder if there will be a counter, that will be a good problem for a qualitate because that means you sort of reach market adoption. But I wonder if there'll be a counterpunch sort of dynamic where hedge funds are thinking about qualitate more as setting the bar and then having to find some differential view of why that bar has some fallacy embedded.
00:52:17 · Brett Caughran, Sagar Kadakia
[00:52:17] S1: And so I wonder, you know, it's like in with with Yipit, for example, there was a sort of first order chapter, and now it's a second order game where it's like Yipit says the data's good, top line's good, but it's because they're discounting, so they're gonna beat on top line and miss on gross margins, and everyone's positioned long, so it's actually a short, right? Yeah, like I wonder if there will be a counter, that will be a good problem for a qualitate because that means you sort of reach market adoption. But I wonder if there'll be a counterpunch sort of dynamic where hedge funds are thinking about qualitate more as setting the bar and then having to find some differential view of why that bar has some fallacy embedded. [00:52:58] S0: I I think uh, and Yipit's a great business. I think uh the difference with Yipit is Yipit is presenting the same information to everyone. [00:53:05] S1: Yeah, right.
00:53:05 · Brett Caughran, Sagar Kadakia
[00:53:05] S1: Yeah, right. [00:53:06] S0: The difference on qualitate is every fund will have a slightly nuanced set of even on the same company. Let's let's just use Microsoft or Nike as an example. Uh a hundred funds on uh the same day and the same time. So they all have the same infrastructure now. So we're 12, 24 months into the future. They all launch at the same exact time. They will have to have slightly different questions, right? It's unlikely that all of the audio questions are gonna be identical, right, to each other, uh, or all the questions they want. And so therefore, uh uh all the responses are gonna be slightly different, right? Uh and the responses won't just be slightly different because the AI interviewer is dynamic, it could be vastly different uh uh on how the interview works. And so I think um part of it is you will always have an edge to the extent that you can ask the best questions and to the extent you have the right infrastructure to capture primary intelligence the fastest. Those two things to me are if you can solve those problems, and I already see with funds trying to solve those problems, how do we ask the best questions and how do we generate primary research the fastest? That will be your alpha.
00:54:13 · Brett Caughran
[00:54:13] S1: Yeah. Have um have you thought about uh corporate access at all? The sort of the, you know, people's sort of logic through all these things, see AI doing more and more. It's like, well, we still have the relationship with uh, you know, corporate IR, corporate CFO, but it's not that big of a leap to start um sort of, you know, um calibrating these voice agents to go have a conversation with an IR, which probably means the IR has their own voice agent. They're uploading the context that they want to disclose, which by the way, sort of brings it back to the compliance consideration. Like that's a pretty good, that's a pretty compliant situation where investor agent speaks to IR agent and all the little things that IR is spending time hand-holding investors through. There's that's probably a more from first principle is a much more effective way to handle that customer relationship. Now there's body language and relationship and all the sort of things beyond that. But uh curious your thoughts, thoughts there.