00:21:39 · Ying Hua
[00:21:39] S0: I would probably say that I definitely think we're biased, right? If everyone can use Cloud Code to build every process, then I think we don't really need to exist. So I'll probably say I do think my answer is biased. Don't think it fully works and I'll tell you why from a few angles. Number one is data pipeline. So you need to schedule build it right. Something is easy you run it every day every whatever. Sometimes it's not. Some data comes out of random interval. How do you figure out when to scrape? All right that's number one. Number two is you need to store that somewhere. You're not going to run everything at like all at once right. Like you need to store that data somewhere. Number three, if anyone has done scraping manually, you will know that website changes, whatever deterministic thing is going break, pretty much for certain. In the span of a year, every single scrape you build probably has some changes. How do you adapt to that, Right, and then lastly I always say I'll tell you this and this is I think, I don't think it's fully solved but when we first started out implied, we actually wanted to do quantitative stuff meaning like natural language quantitative stuff and we quickly basically decided that language model wasn't good at it. And even though language model has improved, I don't think some of the problems have been solved. I'll tell you why. AI generates Python code. How do you verify something? How do you verify the logic you're building there is correct for a non technical person who doesn't cannot read Python code? It's like we think AI can help. I don't think it solves completely. And the reason why that's important is I remember very interestingly when I was at Citadel, we had, data scientists, so they put data scientists onto our team. And we would describe something and they'll come back with this is relevant, this is not relevant. And every time I was like I need to see your code. And I picked up Python when I was at Citadel but I need to read the code to understand what you wrote. And then I can tell you oh you made a logic flaw here. It could be just like one sign difference. It can be very small. And it's just really hard for non technical people to truly evaluate that. I think some of ourselves, like the way we approach it, is you might not know the code but you should know what the right answer is.
00:23:57 · Ying Hua, Brett Caughran
So you can evaluate if the answer is good or bad based on certain points like kind of sanity check. It's kind of like if your intern hands you excel it's not you can check every cell you're going sanity check a few slots. No way to do that. But I think, [00:24:12] S1: a [00:24:13] S0: lot of especially the more complex things I think is really tough. So those two things I think is really really tough. It's like I think clock co building something is very good for one off things. It's harder for a continuous production level self improving, like that part is difficult. And I think going back to just from an industry perspective, I don't think like this is actually true for people's process today. Think a lot of the work that individual team has done is duplicative work. Meaning think about company release earnings. How many Amazon release earnings? How many teams right now are updating their model? There should be one single source of truth. It should not be everyone doing their own thing. Same thing with data pipelines. Like I don't think it makes sense for everyone to try to, you know, I cover 50 stocks, have 50 data sources to manage all those 50 data sources with a pipeline through. Would I rather just use a company that can take care of all that? And how I analyze it is up to me, but how the data is gathered should be just one company. But that's another way to think about it.
00:25:23 · Brett Caughran, Ying Hua
[00:25:23] S1: So we're sort of at a moment in time where in horizontal versus vertical debate, labs sort of are ascendant, in particular Claude is ascendant. And to Kay's point, many of our clients are sort of building dashboards with Claude code. But this sort of topic of agent debt and this topic of tokenomics sort of growing more How in do you sort of see these things? Because it is seductive to sort of spin up a dashboard and sort of think that it's institutional grade, although most sort of deep AI vendors like yourselves we talk to sort of cite that as a sort of not great decisions. How would you sort of frame those two buckets? And what do you think is next for clients in terms of using and building dashboards? [00:26:15] S0: Yeah, I never would say dashboards back. I think think, know, I think, Brett, you've seen our product, right? We're one of [00:26:23] S1: the
00:26:23 · Ying Hua
[00:26:23] S0: few vendor out there that actually have a dashboard. We actually think there are certain things that you shouldn't every not every answer should be a chat answer right. Some information should be serviced to you without you having to ask it should just be there. A reason why people like to stare at a Bloomberg terminal right? Like there's value in that. So I would never say dashboard is bad. But two things. Number one is I think the idea of a static dashboard is bad even for the same company, even for the same industry. Like what the dashboard should look like today versus what you should look like two months from now for next earning season could be different right. This is based on domain knowledge things. So one is how often do you want to be changing that dashboard? Do you want do analysts and PM want to spend their time deciding how to change? Or can you use a platform like ours to be like you have all the data already, I just want you to present data in different ways or I wanted to reorder data in different ways so you can show me what's important what's not important. So that's number one is just like how much time you want to do it. Second thing is, I do think that the ultimate form of what investing should look like, the process, is not going to be mainly dashboard driven. So what do mean by that? I'll give you an example. So I used to love chart books. I don't know if you guys like look at chart books, like you know you do certain charts, whether it's like an estimate revision stock, whatever it is, and then you have it, you know, do a PDF, you flip it through, right? You get a lot of information, you get a lot intuition from it, it's great. But once you get to like chart book number six, you're never actually looking through the chart books, right? You're not actually going through okay what's new, what's worthwhile, right? So dashboard, I think there is going to be a similar dynamic. You're going to build a lot of dashboard. Everything is going be very cool. At some point, you're going have too many dashboards to manually check. So what do we think is the ultimate end of, like, investing with AI? I think for us, it's like you should be able to create dashboard or however many view. The dashboard is just like how would I want to view this particular data source, right? That's all it is. It just everyone has a different preference.
00:28:31 · Ying Hua
What I think is important, what I'm screaming for, what you should be able to create dashboard very easily, Create any metric, any visual very easy. That's number one. But two, you should have someone reading through it and tell you. It shouldn't just be building dashboards. It shouldn't stop there. It should be hey Brett, basically you have a dashboard that tracks 100 different data points. I'm going to check these 100 data points for you to tell you hey Brett you should look at this one, this one's interesting, this one has a big jump, this whatever. Like my view is dashboard is a means to an end for you to reach some conclusion. And the purpose of AI is should be treating like you have an intern who read those dashboards for you, who build that dashboard for you, read dashboard for you, filter all things that's important, and alert back to you so that you don't have to have to mentally spend time thinking through those. Makes a lot
00:29:22 · Brett Caughran
[00:29:22] S1: of sense to me. I mean, I think one of the fundamental challenges of office software, investing software, has been this deep heterogeneity of investment process in sort of by style and by coverage area. Right? Yeah. And so really, like the only scaled software has been like the terminals, which are just sort of like an aggregation of data and allow for this customization in a launchpad, right? Like all of our launchpads look differently. Yeah. How do you think we take that same concept? Like how do mass market scale to this degree when you're sort of building software? How do you allow for that mass customization of dashboards to your clients?