August Kiles: So our goal is to be kind of a one stop shop for anybody working in capital, growth equity, private equity, family offices, that sort of thing. And so we help you track deals for the whole life cycle of your fund.Intro: Built this week, breaking it down. Built this week, we show you how. A fresh idea, a clever tweak, you locked in shoe. Built this week.Sam Nadler: Hey, everyone, and welcome to Built this week, the podcast where we share what we're building, how we're building it, and what it means for the world of AI and startups. I'm Sam Navler, cofounder here at Rise Labs, and each and every week, I'm joined by my friend, business partner, cohost, Jordan Metzner. How's it going, Jordan?Jordan Metzner: Hey, Sam. How's it going? Happy happy week. Excited new week in AI. It's been a crazy new week with new models and lots going on, and great new guests as well.Sam Nadler: Great. And this week, we have a special guest, August, from Emblem. August, give us a quick intro. Hey, everyone.August Kiles: First of all, Jordan, Sam, super excited to be here. Thanks for having me on. My name is August Kiles. I'm head of product at Emblem, which is a startup that optimizes workflows for investment funds.Sam Nadler: Amazing. And before we get into the docket for this week, please don't forget to like and subscribe. We've just crossed 20,000 subscribers on YouTube. We have episodes out each and every Friday, so hit that bell, hit that like button. And the docket for this week, we're gonna go over a little portfolio scenario tool we built with Emblem in mind.Then we're gonna jump into the Emblem product, and finally, we're gonna cover the latest and greatest in AI news. So with that in mind, you know, August and I discussed a few different potential future features for Emblem. And just to keep in mind, as August mentioned, Emblem is this and now we'll get into it a little bit later, but it's this amazing tool for managers and investors to, you know, better access their data, build financial reports at the touch of a button, you know, discover their alpha within seconds. We'll get into it. But one of the tools we, you know, features we discussed was potentially building a portfolio scenario tool, which is what I have displayed here.So assume our portfolio has 30, you know, seed companies. We have Construct, that's a great name, SwiftPay, AgroLogistix, which we invested as in the seed round in 2022. But there's potential macro scenarios on on the horizon. Maybe regulatory pressure increase. You know, we can we can increase the the speed of change.Maybe it's it's only a temporary potential macro scenario happening. Markets sentiment will will leave relatively unchanged, and the time horizon, six months, twelve months, twenty four months, whatever, run the simulation, and we can see how this potential macro scenario could theoretically affect our portfolio. AI is doing its work, beep, bop, boop, beep, bop, and boom. And here it is. So from regulatory friction point of view, you have your, you know, series of companies that, you know, look like they wouldn't be affected.You know, some could be in the advantage zone, some could be in a potential stress zone, and some I could keep your eye on. So this is obviously, you know, a demo tool. This is something maybe Emblem could layer on in the future. I wanna show one more scenario. Liquidity supernova.Jordan Metzner: Every company hasSam Nadler: a liquidity event. We've got cash flowing. Everything everything is on the up and up, has been greatest investors of all time over here. But now, anyway, August, time to transition. Tell us about Emblem, what you guys are building, who the customer is, and how you're using AI.August Kiles: Yeah. So I think our landing page here says it really succinctly, the last platform investors will ever need. So our goal is to be kind of a one stop shop for anybody working in capital, growth equity, private equity, family offices, that sort of thing. And so we help you track deals for the whole life cycle of your fund. Starting the second that you get a data room in the door, you sync it with Emblem, we index all your files, and then there's a whole host of different features that allow people to just speed up the things they do and do them more accurately and better to free up your time for other things.So in the diligence process, we have great search tools, you know, that dive into single deals and also compare, multiple deals, anything that's ever been uploaded into Emblem. We have report generation that can automatically export into Word, PowerPoint, and Excel. We help build financial models. We accurately extract, you know, all the financial data from your data rooms and everything like that. And there's a whole host of UI features built on top of all of that to make the, you know, ease of use as, you know, significant as possible.So that's kind of a general overview. And then the cool thing is that, you know, the space of the funds investing in different sectors and different stages, there's so many different things that people need. And so, you know, what's really cool about something like what you guys just built is it's something that could be useful to, you know, definitely like a large portion of our, you know, potential customers out there. And there's kind of this constant, you know, flywheel of we need to keep building the next thing in order to keep winning over these markets because there is such variety in, you know, the workflows that people are doing and, you know, what they personally can benefit from, and it's so easy to build new features these days that we could ship stuff kind of on demand for people as needed.Jordan Metzner: That's really awesome. So, tell us just a little bit of like, I mean, obviously the whole product is based on AI, so I presume like without AI, the product wouldn't even exist maybe.August Kiles: 100%, yeah.Jordan Metzner: So it's almost like, you know, what could you do with AI that you couldn't do before? But I guess, like, literally everything here, right? I mean, like So, I guess, maybe take us from, like, the old way to the new way using Emblem. I guess, previously, like, all these were stored in, I guess, like, separate silos of datasets. Right?August Kiles: Yeah. Absolutely. So I think first, I'll just, you know, speak very briefly on the founding of Emblem, our our founder and CEO, Luke Ashrambrand, and the second that LLM started to become, you know, more widespread in in '20 late twenty twenty two, early twenty twenty three, he'd been working in private equity. He'd done an internship in venture capital, and he immediately saw the potential for this technology to, you know, really play in the space. And so, I'll give you know, a few different examples.But in the past, you know, you're a private equity firm, and you're thinking about acquiring a business, and the CEO sends you a data room that has everything you're gonna need to do due diligence, right? The only thing you can do in that situation, you know, pre AI, is have a bunch of associates at your firm just comb through it and look at everything in detail. We're talking about, you know, dozens, if not hundreds of hours of work to get through a data room like that. And with AI, you know, you still have to go to the same level of detail, the same level of granularity, But, know, let's say you need to know something very specific about the team, or you need very specific revenue numbers and you don't necessarily know which spreadsheet they're gonna be in, you just ask Emblem. Hey, you know, where's the revenue for, you know, 2017 for, you know, this particular product in the company or something like that.And in a matter of seconds, you've got it. And then you know, you can also take it a step further with reporting where you can provide Emblem with a whole list of different metrics that you want to extract and what kind of analysis you're looking for. And you know, once again, matter of seconds, you have the results. So rather than have, you know, one associate looking, you know, for hours to, you know, make sure they've gathered every data point and then writing up different summaries and that, you know, that all of that. You can just have Emblem do the same thing.The associates still do the last 5% and, know, figure out that their job is to figure out what to ask Emblem and to then take that information and present it as needed internally in their organization. But so much of the heavy lifting in between there can be automated by AI. Yeah. Do you would you like, aSam Nadler: large fund, I mean, would they even need nearly as many analysts using Emblem as they theoretically would have needed before?August Kiles: I think that's always the interesting question, right? And in general, I tend to come out on the side of it just frees up the analysts to do more, rather than you need less of them. I, yeah, exactly. So I don't really see this being a product that takes entry level jobs in the industry away, I think it definitely one of the things that our potential customers tend to really like early on in the sales process is we talk about getting to a no faster. So with Emblem, they can rule out investments way faster than before.Right? Oftentimes the investments they are gonna do might take just as long because they wanna be extremely detail oriented regardless if they're gonna actually put money in. But if you can say no in, you know, say just a few hours of work instead of a hundred hours of work, then you can look at that many more companies. Right? And and maybe you find a few more diamonds in the rough that you wouldn't have even looked at before because you didn't have the firepower to look at as many deals.So I I think that's more where this really shines, if that makes sense. And I I don't think that ultimately takes jobs or, you know, results in these funds needing less people. So if you look at the the the way that their compensation is structured with their management fees, you know, the low level employees are not very expensive for them. It's really not moving the needle on the fund returns. What definitely moves the needle is finding the right deals.Sam Nadler: Yeah. And access to more deal flow, as you mentioned. With this tool, they could theoretically increase deal throughput or, you know, nose faster. Exactly.Jordan Metzner: Yeah. Cool. And is there a part of the market where, like, this specializes in? Sam mentioned, like, large funds, like, is it better for VC? Is it better for family offices?Is it, like, better for private equity? Is it across the board? Kind of Where is kind of the bestAugust Kiles: use It's definitely useful across the board. The way we see it is that as the stage progresses from, you know, seed stage all the way to private equity, the utility of Emblem increases because the complexity of the data rooms increases. Right, I mean you guys know firsthand like seed stage investing, there's a lot of instinct, a lot of intuition, a lot of vibes, right. And the financial models are not as complex, and there's less information. So all of that you know, means that it takes less work for a human to analyze the company.Whereas if you're in like the private equity, you know, range, and you're getting in these data rooms that have thousands of files and extremely complex financial models with years of data, and you know, potentially big companies with lots of different verticals and lots of different products, you know, seemingly multiple orders of magnitude, higher complexity. And so we think that's like the ideal use case, but we do see itJordan Metzner: as being useful across the board. That's awesome. Tell us just a little bit about of like what kind of models and what you guys are using. I mean, I'm presumably using all the different labs and different things across different laboratories, but maybe just, like, you know, I know we're gonna jump into the news in a little bit, but just maybe the impact of some of the latest stuff from Anthropic and OpenAI and the impact on on just, like, maybe even your development, as well as kind of the product itself.August Kiles: Totally. Well yeah, I think your assumption is correct, we do use them all. And you know, we have a, the kind of your classic like multi agent structure, we are good at knowing which model to ask for which task and which question and that sort of thing. And you know, I think we probably use Claude and Gemini the most in our actual products, but we we do use all, use, you know, Grok, we use OpenAI, all of it. In terms of with our own workflows, we'll get to this more with the news also, but we've definitely watched the models play an ever increasing role in our software development.I mean, we had a bit of like a jump start last year around this time where we started relying on them maybe a little bit more than we should've because there was some over optimism in terms of how well that, you know, they could write code at the time. Had to kinda roll that back a little bit over the summer. We're having too many bugs in, you know, code that we had pushed, you know, with a little bit too much urgency. So prod, I mean, it's just kind of growing pains for a startup at the time. You know, we were doing a huge refactor of the platform as it existed, and we're hoping to as soon as possible.So we spent a lot of the rest of 2025 kind of not relying on AI as much. But with the recent updates, it's kind of it shifted back towards being a majority of the code that is written for Emblem is AI generated in some capacity. We can touch on that more with the news. And then also the new models have really improved certain things in the platform, specifically financial modeling. So, you know, like the new Opus is really amazing at generating a financial model and outputting it in a spreadsheet and that sort of thing.And they we've really seen a huge bump in performance thanks to that, which is really cool to see.Jordan Metzner: That's awesome. Yeah. That's like a step function improvement, getting the models to work with old file types, in this in this case, like, you know, Excel models. Because obviously, it's just gonna get a lot better from here. Maybe toAugust Kiles: the Exactly.Jordan Metzner: Excel gets completely abstracted, like the database has or whatnot. Okay. Super cool. Should we jump into the news, Sam?Sam Nadler: Yeah. August, thank you for giving us, you know, the the high level on Emblem. If anyone's interested in checking out Emblem, go to emblem.pe, and let's jump into the news. We Jordan's alma mater is Amazon, so we we cover it quite a bit. And we also love Claude Code here, so we thought this would be a fun story.But seems like Amazon engineers are, you know, clamoring to get their hands on Claude code versus the internal tool, which I think is key role. So, you know, Jordan, just as a future Amazon or as a past Amazonian, what do you Yeah.Jordan Metzner: So I have a few friends that work at Amazon, and I remember talking to them a few months ago, and they were really behind on using LLMs for coding. You know, they were, like, three or four OpenAI models behind. I don't think they had any anthropic models. And then I just talked to some a few friends recently, and now they said they're kind of full stack on this key row. But one one friend told me that in the background, it's just an older version of Opus.So it's just a wrapper on a wrapper, which isn't surprising. Right? Like, I mean, obviously, these companies are worried about security compliance and, like, regulations and a bunch of other things. Same time, wanna kinda push their developers to having the best tools. But you see, developers are kind of, like, going crazy because there's such a step function improvement between Opus four point five and four point six that, you know, developers wanna get their hands on it as soon as possible.So it does seem that, like, at least at the, you know, at the FANG kind of, you know, MAG seven companies, the gap from launching new AI models to the market to, like, getting them implemented at the enterprise level is is getting shorter and shorter. And that obviously is, you know, dangerous for startups and good for, you know, probably the mag seven. But, you know, as they get closer and closer to be able to launch, like, these new models when they come out, then they become, you know, more and more powerful rather than, you know, having to use kind of old technology. I think, like, you know, we're lucky in that sense because, like, as soon as the new models come out, we're able to adopt them pretty rapidly, you know, even same day. I don't know how you guys feel about that, August.But, yeah, I think it just shows that this time is collapsing. Maybe they're not there yet, but they're gonna be there really soon where, like, new model launches, even maybe we'll see that they get it first. I mean, I think I saw yesterday, OpenAI announced that everyone at NVIDIA is getting Codecs, so you can just just imagine that, you know, like a new version of Codecs is ready, and maybe they launch it to their clients first, so that, you know, the whole model kind of flips, and these enterprises are getting access to the models even before the public does.August Kiles: No, I think you're totally right. I I think we will see the enterprises getting first access sooner rather than later. You know, especially like I would imagine with Gemini, it'll be these new model, the new versions will be used in house at Google, you know, long before they're used anywhere else. We already see with Anthropic, right, they're using Claude code heavily internally, presumably with new versions of Opus, you know, pre release. I guess I have a few thoughts on this article.It was interesting to hear, you know, kind of that insider info that they're behind on which model of Opus they've got running. Because, I think that's really significant because the latest release was truly an inflection point in my opinion. And I have found in my own work that, you know, especially if I really do a good job, because I do write a lot of code as head of product at least at this early stage of the company, but if I if I do a great job with the architecture and the prompting, it's actually pretty rare that I need to make any adjustments to what, you know, the models end up creating, and that's a lot different than it was a year ago or even four months ago. So if they're leaving that on the table, that seems like something that's going to have to be corrected fast. My one contrarian viewpoint on this is that I don't really think that, I think people tend to like bandwagon these different products like Claude Code.At the end of the day, everything out there is really good and using, you know, say Cursor or Copilot with Opus 4.6 is going to be so much better than using Claude Code with Opus 4.5 that it's not even funny. I personally do still use cursor sometimes. I really like certain things it can do. You know, you can launch the browser inside there and it can For making UI tweaks, that's really helpful. The debug mode is really helpful cause it'll tell you, okay, go and click on this, and then it automates the logging and everything.So I think that all of these tools are great, and that the common thread is simply how good are the models behind them. I think it's the same thing, know, in terms of say Emblem and our competitors, right, we're all using the same models. We all good rag systems. So a lot of it is truly the product itself and the user interface and how quickly and how well can we build exactly what our users want. Because, you know, our competitor with the new version of Opus is always gonna be better, you know, on the back end than us with the old version of Opus, and vice versa.Jordan Metzner: Yeah, totally. I completely agree with you. I think it's like a new regimen jumping up to 4.6, and hopefully we'll keep seeing these step function improvements that kind of make things get a lot better. It's obvious that like the next step is kind of these like swarms or teams and kind of managing these teams of teams, and you know, I've been playing with it a little bit. It seems still a little rudimentary in like how successful they work, and at least on the Claude side to me, they've they haven't been able to end to end kind of eliminate major issues that I have to kinda go back and clean up or whatnot.But, yeah, I found the Codex models to be really good. And I think that, you know, they announced this new frontier tool that they did launch to enterprises and haven't launched it to the public. So, you know, maybe it's already happening and we're kinda behind the curve there. But, yeah, some great insights. Sam, what about you?And I know you're not even using Cursor anymore.Sam Nadler: No. But I agree with August. I think, like, whether it's it doesn't matter what IDE you're using or or not. I think that the model there has been such a change in my workflow. I'm not a developer by trade, August, and, like, honestly, lately, I've been, like, pushing code.The pace of development internally over the past, what, six weeks, eight weeks has just like really increased. But it's all because of a 4.6. Yeah. Let's transition just in the for time. So Musk, big reorg at X AI.Obviously, was the SpaceX merger announcement, I think it was last week. Any high level thoughts here?August Kiles: Yeah, I guess, you know, the first thing that comes to mind here is it's always, you know, he's always five steps ahead of everybody else in, you know, everything he's doing. So my first thought is that there's definitely kind of a longer horizon thought process going on here. And I was actually just listening to a podcast yesterday that he was on, and you know, keeps, he talks about the digital optimists, right, to go along with the physical optimist robots that he's building. And so I do wonder if part of this is him saying, well, we're not as interested in being an API that, you know, companies like Emblem rely on for, you know, in his mind, like more basic workflows that are completely uninteresting when you're thinking about going to Mars. And he's more thinking, I actually want this company to build the kind of full stack like digital human that can do any sort of work that any human would do end to end.My guess is that's what this is about, and he's kind of hunkering down and doubling down on that. But I'm curious what you guys think.Jordan Metzner: That's an interesting take. I mean, you know, let's not forget that XAI and now SpaceX have merged, which is separate from Tesla, which has the Optimus robot. So still separate companies, you know, who knows where that'll go in the future. You know, I think if you look back to Elon's tweet a few weeks ago, he he referenced how good the Anthropic models are at coding, and I think he thought that they would have a better model for coding than they do today. And I think historically, when Elon not happy, like Elon pound table and like everybody has to go and change it up, you know.But I think it's more so, and I think this is kind of like an interesting thing that I was reading in the Matt Schumer blog post just about like one of the reasons why coding is the first solution of like these LLMs is because then code can write more code, to write more code, to write more applications. Right? So I think if that's the case, even internally at xAI, and they just like we mentioned earlier about Amazon, if they're developers if you're a developer at xAI, and you're demanding your boss that you wanna use Claude to code, which why would you not, considering it's the best Right. Know, Claude or Codex to code, and you're not saying, hey, give me x AI's coding tool. Well, think that, like, in and of itself is the sign.Right? And, you know, you mentioned earlier you guys use all the models. I mean, I can tell you honestly, like, we don't have Brock models in production. We use we definitely use, like, frontier models, like, you know, Anthropic and and OpenAI and Gemini, but we don't really use the Brock models in production. We haven't really found them of, like, high quality of use case yet.Maybe I'm wrong, maybe we're like not, we're missing the boat here, but I think the reality is like, if the internal developers at xAI are not demanding to use xAI, that that's enough to probably make a change in management. I know at Facebook, like, wanna use Claw too, I mean, I think it's it's happening all over the place.August Kiles: Right. Well, I think it's an interesting leveling of the playing field, right? Because Anthropic puts out the best model, and you know, if you're willing to kinda swallow your pride a little bit, if you're working at one of these other labs, then, you know, you can then use that model internally to stay competitive on building the next one. So I'm I'm actually kind of, your your point is interesting and it makes me curious as to, you know, to what extent they're just allowing people to use Claude code if they want to versus saying, no, you have to use, you know, what we've got here. And also truthfully, we don't really use XAI that much either.It's mainly actually, there's certain features where we provide the user with a dropdown where they can select which one they want because sometimes people already have a preference, and that's where we have xAI in there, is just in case the user wants to use it, but all the stuff that we're doing on the backend, we don't use it.Jordan Metzner: Yeah. Yeah, totally agree. Okay, cool. Well, August, this was an awesome episode. Great to chat AI, great to chat Emblem, great to meet, and thanks for joining us today.Sam Nadler: Yeah. Episode 31, August, thank you so much. Don't forget to like and subscribe, and episodes are out every Friday. Thank you, everyone.August Kiles: Sweet. Thanks, guys. I this was great. I appreciate it.Jordan Metzner: Thanks, August. Thanks, everyone. See Bye you bye.
.png)
