00:24:00 · Khe Hy
[00:24:00] S1: been finding? Are your clients finding it easy to, like creating a skill is relatively easy, but it's imagining what skills you need and how they might fit together. Like the kind what we've been discussing before, art of the possible. How has skill building evolved or shown up in your work?
00:24:24 · Brett Caughran
[00:24:24] S0: Yeah, it's constantly changing like everything, right? And I'll give you sort of one example. I had a deep dive skill that created 100 page primer. And you have to sort of the prior model iterations, you sort of go through like, hey, these are the 12 analyses I want to do on this company, right? It would run the same 12 analyses on every company. It would adhere to those skills. One of those analyses may be balance sheet and liquidity, which could be important for a leveraged company, but not that important for a company that has net cash. But in sifting through that primer, I'm still reading those sections for every company. So there's sort of like the concision dimension was just not that satisfying, right? This is one of the things that unlocked with Fable and Soul is sort of ability to sort of rethink that operating system. One shift I've made to that sort of simple primer skill is this concept of, I call it twenty-sixty-twenty, where it's like, it's 20 pages of standardized sort of business overview, setup overview. But then like a good senior analyst, that front end is actually trying to understand what the key debates are in this situation and spending the meat of the primer actually going in and doing analyses, right? And so it's identifying what matters in the first 20 pages, and then it's going to give like a deep, you know, custom bespoke analysis of each of those factors. And this has been like a complete unlock for me terms of the quality of the primer that you can create with this sort of skill. But it also creates a totally different system. Have to create the pool of deep of key driver workflows that the digital analyst can call from. So it's like, I'm kind of like decommissioning the old structure to create a new structure that that really leans on the new capability, the sort of more end to end capability. The exciting thing is you're getting to the point with Excel fluency. These new models are much more Excel fluent, Where after you've identified the key driver, say looking at Brunswick that has a boat business and I want to spin up a price like an aftermarket pricing, secondhand pricing tracker to go scrape boat trader, etcetera, you can start to sort of build into the system, Hey, this aftermarket boat pricing is a key driver because it sort of impacts a price gap of new, etcetera.
00:27:00 · Brett Caughran, Khe Hy
Do you want me to sort of create that for you? Right? So you're almost like spoon feeding. You're creating this system for people that to me is very close to, you know, something that's going to wow a lot. Like all they need to do is sit in the cockpit and almost you just have to press the right buttons rather than, you know, think through some complex prompt that we had to do, you know, twelve months ago. So that's I think what the that's that's what the the new models have unlocked. That's what the connectivity, we can talk a little bit about connectivity systems have a lock, the sort of ability to more seamlessly integrate MCP structures, but also the headless RMS movement, which is so exciting because I've always hated RMS systems so much. It's so clunky. That to me is sort of connectivity overcomes one of the biggest fundamental issues that AI systems have had for investors, and that's the sort of context quality. Yeah. [00:28:03] S1: So let me on that context point, it seems like and you're closer to this than me it seems like the MCP universe has completely expanded in the past six months. Right? Every day someone emails me about a new MCP, whether it's for expert calls or investor transcripts or, you know, we're talking about BlueMatrix or Aeriea, I always pronounce that one wrong, FactSet and so on. I'd love for you because I don't know this world that well. I'd love for you to kind of give a couple minutes state of play of the MCP landscape for investment firms.
00:28:03 · Khe Hy, Brett Caughran
[00:28:03] S1: So let me on that context point, it seems like and you're closer to this than me it seems like the MCP universe has completely expanded in the past six months. Right? Every day someone emails me about a new MCP, whether it's for expert calls or investor transcripts or, you know, we're talking about BlueMatrix or Aeriea, I always pronounce that one wrong, FactSet and so on. I'd love for you because I don't know this world that well. I'd love for you to kind of give a couple minutes state of play of the MCP landscape for investment firms. [00:28:46] S0: Yeah. Yeah. And I'll sort of tell it with the FactSet stock chart up on the left terminal. My left screen, you know, the FactSet stock went from sort of the high 400s to the low 200s during the Claude kerfuffle. [00:29:11] S1: Almost 50 down.
00:29:13 · Brett Caughran
[00:29:13] S0: Almost 50% down, even though the fundamentals have sort of been pretty good. And I think, you know, sort of the principal concern has been the shifting interface for investors and deal makers. And I'm a FactSet subscriber. I even noticed that. Like I very rarely log into the FactSet terminal anymore. So this is sort of the setup, I think, all vendors of the platform is burning, right? But what's the sort of next thing? Now, the FactSet MCP is really good actually. And it's very sort of critical in a number of the workflows that I'm building to make them institutional grade. Because if you go to chat GPT or Claude out of the box, and you ask for a primer on a company, it is going to pull principally from the retail aggregators, the Zacks and Motley Fools and Seeking Alpha, Alpha. Like no shade to all of those. They're fine businesses, but it's not institutional grade quality for hedge fund quality research. And I think we all know like the training corpus of the foundation models themselves are highly influenced by Reddit and Wikipedia and blogs, etcetera. So it's like two layers of low quality context for, you know, even if the reasoning engine of the LLM itself is strong, you're sort of, know, the underlying quality. And so improving the quality of the data has been critical. That has been very hard to do for the first two or three years of AI. Firms had to create data lakes, whether it's Snowflake or Databricks and, you know, API heavy systems. And it was very difficult, very difficult to do, to sort of ingest, clean the data, expensive to do. The APIs were very That has been rapidly abstracted as, you know, I think the whole ecosystem has seen this trajectory towards MCP, which sort of, you know, Tom Lee in our session explained is like an MCP with like good instructions upfront, or sort of an API with good instructions upfront. But the MCP ecosystem has really been expanding. You can start to get a very good baseline level of institutional grade research in MCP stack. Now, it's not inexpensive. We could talk about that. I think that's a problem that needs to be solved. So it's sort of layer one. Second layer
00:31:46 · Khe Hy
[00:31:46] S1: for equity research, it's coming mostly from FactSet and then the aggregators like BlueMatrix and Era. Is it Era or Era?