A Product Market Fit Show | Startup Podcast for Founders
00:32:01 · Pablo Srugo, Chaz Englander
[00:32:01] S1: And yeah, that's how it started. It's, know, very similar to the kind of like Toby Shopify story, right, where he builds like this this ecommerce store so he can sell snowboards and like the snowboards are okay, but the store everybody everybody wants. [00:32:14] S0: Exactly. Because the thing is as well is like where it was and to a certain extent where it is, it's a productivity tool. So it wasn't necessarily making us better investors but it was making us more productive. [00:32:25] S1: What did it do? That was my main question, yeah, like what exactly was the product?
00:32:29 · Chaz Englander
[00:32:29] S0: It was basically, we would receive an opportunity. So when I say an opportunity, anything. It could be to invest in a startup, it could be to invest in a fund, it could be like some random co investment, it could be like a bunch of like stocks, it could be anything, right? And we'd get these opportunities through a whole variety of different avenues, we get them via email, and like there's a bunch of things that you do really really quickly as a human to validate those things. Now almost all of that information that you do to validate is not in the materials you received, because what you've received is a sales pitch, sales materials. They're not gonna tell you the bad things or the things that you might quickly get you to an answer early on, right? They're just trying to get you to like pick up the phone or get on a call or potentially be interested. So we basically set up this... Well, then we were just referring to it as like automatically calling functions. Now you would call an agent, but it would receive an email, it would look at the opportunity, and it would go and then do a bunch of things. So for example, if you... And maybe relatable to this audience, If you receive an opportunity to invest in a startup, right, one of the very, very first things that you do is you look at who is running the startup and you might see whether you've got like any connections in common with that particular individual on something like LinkedIn. Right? And you can get to the answer programmatically. So an example might be, I receive a deck and if that's one of the first thing that I do, I don't wanna look at the deck. I just wanna ping Joe Bloggs on, you know, WhatsApp, hey, just got this opportunity. What do you think of, you know, James Smith? Right? Or whatever it is. And so it was just building automations off the back of what you'd receive that could enrich the data that you'd ultimately got via email. So either via different data points, the example of, like, connections in common, things in the news. Now if it was a consumer company, you may have loads of bad reviews on Trustpilot or equivalent. Or if it's a a fund or or a new GP, are they posting strong political views on Twitter? I don't know. Like, you know, there was a lots of things that you could do. So that's where it started, and it just got more sophisticated.
00:34:35 · Chaz Englander, Pablo Srugo
So it wouldn't just give you the bullet points, it would actually create you like a word doc or a PowerPoint presentation of the opportunity that you could really quickly like get ramped up on and then quickly roll it in or out. So that's where it started. And for better or for worse, I mean, that's kind of what the product does today. We call it a workflow automation product for finance, but basically it's doing that, predominantly being used for the automation of the creation of materials. So Word, PowerPoint, Excel, etcetera, and particularly things that folks in finance are doing on a repetitive basis. [00:35:08] S1: So it's kinda like, I mean, like data enrichment, I guess, is is the primary piece. When you mentioned the the WhatsApp part of it, does it actually go out and reach out to people for you, or it just tells you like, hey, you know these people, maybe you wanna reach out, and then you can you can do it on your own?
00:35:08 · Pablo Srugo, Chaz Englander
[00:35:08] S1: So it's kinda like, I mean, like data enrichment, I guess, is is the primary piece. When you mentioned the the WhatsApp part of it, does it actually go out and reach out to people for you, or it just tells you like, hey, you know these people, maybe you wanna reach out, and then you can you can do it on your own? [00:35:21] S0: I think it's heading that way. So being quite clear how the product has evolved and what it is we do. So we describe ourselves as an AI workflow product for financial services. In other words, an AI workflow builder specifically for financial services. Now what that actually means in practice is number one, well, is financial services? Our customers are predominantly, you know, banks, very large consultancies, we include kind of big four in that, and then and then asset managers. But it's very much an enterprise product. Enterprise, I mean, all the way to the extent now we, partly due to the demand, partly due the way that our business model is evolving, know, we don't really look at customers that are gonna be spending any less than maybe a $250,000 a year with us. Gotcha. Just based on what we do. Now again, when we say what we're good at, it's really material creation and verification. The verification one's interesting, so it's, you know, if you already have an output, you would go and verify that before sending it to someone. You can use really, really clever AgenSys systems to do that in an incredibly sophisticated way. [00:36:24] S1: Walk me through that a little bit more, like verifying what exactly?
00:36:27 · Chaz Englander
[00:36:27] S0: So under the hood, there's going a little bit technical is there's three different AgenSys systems. One that goes and gathers all the information that you would need in order to create some kind of output. When I say output, I just think think an email, a Slack message, all the way through to like PowerPoint, Word, Excel outputs. Think very complicated. So don't think two pager, think like 200 page PowerPoint presentation, graphs, tables, charts, logos with like all the Excel backup files, and then verification. Now the reason interesting that the identity systems are built in that way is that's exactly the same way that humans go through their sort of... We call it the cognitive flow. Right? So as a human, when you initiate a task in your head, you gather data either already in your head or you would go to the different things that you can access on your laptop or phone. You would put that together in something, and then you would check it, and then you would go to the next thing. In some cases, send things out as an example. That's what the agentic systems do, right? Where that's being used a lot is the automation of the creation of the material and the verification of the material. So on the creation, if you think about a lot of where the time goes, particularly more junior members of the team at these sorts of companies, is they are spending, these guys, you you're doing fourteen, fifteen, sixteen hour days quite often at these types of firms, and if you study where the time is going, it's not going in like what we would think is like strategy or critical thinking or analysis or anything like that, it's literally going at like creating a PowerPoint presentation or writing an email. In other words, you already have the information, you're just putting it together and repackaging it and sending it out. And so that has been very popular, and I think the timing of us coming to market in terms of the sophistication of the models and everything else was, again, I think quite lucky.
00:38:22 · Pablo Srugo, Chaz Englander
[00:38:22] S1: Tell me about getting that first customer. I mean, you mentioned it it got pulled out of you. Walk me through that story. Like, how did that all come about? [00:38:29] S0: So that was of one of those customers that were like, hey, you should do this full time. They came on as a design partner. [00:38:37] S1: And this was a a what by the way? Don't need the name necessarily, but is it a bank, it's a... What is [00:38:41] S0: it? A small adviser. It's a small investment bank. They came on as a design partner along with another firm quite quickly, and I think that's the way to go. So I would say as a playbook for b to b companies, you almost can't overestimate what a good design partner can do to speed up your iteration cycle. And I think, you know, something that Arne and I have learned, if I would say there's like a couple of things that are in the founder playbook in our head is like perseverance is definitely one, but another thing is just the speed at which you can learn. Like I don't just mean as an individual, I actually kind of mean about your customers, right? And how you can translate that to the product. So you know, when I speak to people, if you've got a design partner that you're speaking to once a week, you're almost certainly gonna fail. Right? If you've got a design partner that you are working from their office and ideally sat next to them, you're almost certainly going to succeed. In in a in a world, by the way, where there's no kind of barrier to entry when it comes to like writing code or or technology in general.