"The Cognitive Revolution"
00:35:26 · Marc Bhargava
[00:35:26] S1: Yeah. Generally, think about it by going in there and shadowing the customer and understanding all the workflows in their business. And so then we add up who are all the folks, what are their workflows, and we map out the hours at a company. Then what we do is we go through those workflows and hours worked and we look at where could AI automate some of these workflows. And that's trying to get us to this 30% of hours worked automation. So 30% of tasks can be automated. And generally, historically, over the last two years, those tasks have fallen into about four or so buckets. One bucket is customer service and customer support and frequently asked questions. And so if you have any business that has a lot of customer support, it's heavy on that. It could be a good potential industry to go after. A second one is the creation of presentations, copy, marketing, emails, businesses that have a lot of that work stream, also a high degree of automation there. A third is if the data tasks are very repetitive. So there's a lot of filling out forms or looking at sheets or an insurance is a good example of this, Checking the box, did this roof cave in? Yes or no? Did they fill in the paperwork? So that's kind of a third bucket. And now fourth, more recently is around logic and reasoning. So again with insurance it's less automating forms now and more underwriting like can we make a suggestion of how to price this insurance coverage policy based on everything we know. And as models are getting better at logic and reasoning and math, you can have more AI tools kind of with a human in the loop giving you kind of a quasi answer to what you might want to do or say. Legal is also a good example of this where now Anthropic and others have created models that are quite good at reasoning and so a lawyer can say, well, know, what cases am I gonna... What questions am I gonna get on my case? And so you can start to do this, like, more human prep. So those are four very broad buckets, but areas where AI, as they stand today between LLMs and agents, can automate a good deal of things in those buckets. So then when we start going through these, you know, 70 industries, we...
00:37:22 · Marc Bhargava, Nathan Labenz
And we break them down and we actually get on the ground and our companies that we've helped incubate are actually out in the field talking to clients, we are mapping out how they spend their time. And then we're overlaying where does that mapping fit with the buckets we've identified have a high degree of automation. And so that then really helps us underwrite kind of what is the opportunity in this space, what is the opportunity if we were to buy this client, And instead of having it as a customer, have it as a company we own that we can automate large portions of, that we can take on more revenue, that we can create more free cash flow from. So that's the strategy we go in and implement. And that's why it's important to have people who know the industry, know these clients, can get us in the door, but also folks that come from these really strong applied AI teams that know what best in class kind of automation looked like at a ramp or at a rippling or at a Figma or a place like that. [00:38:16] S0: And you do that initial mapping even before starting the company? Like, that that was part of the 70 to 10 cut down?
00:38:16 · Nathan Labenz, Marc Bhargava
[00:38:16] S0: And you do that initial mapping even before starting the company? Like, that that was part of the 70 to 10 cut down? [00:38:23] S1: Yeah. So we do it at a certain level, from the 70 to 10 cut down. And then once we've actually given the first round of funding, in order to get the second round of funding from us, then the team needs to go out and really get the data. And so they get the really granular data from their customers, from their pilot clients, then they come back to us, present that data, then we give them the capital to actually go buy a target, oftentimes one they're already working with. So we do a first round of analysis in mapping it and looking at the industries. As we start placing bets, then, of course, the team has an even higher bar to build an AI native software product that is very useful and to go and win an initial set of clients and then to come back to us and prove to us the automation with those clients. [00:39:05] S0: It's obvious why the owners of these businesses would be interested in this. How did the staffs at these companies react to this sort of transformation? What what have you seen there in terms of excitement, fear, you know, rejection, you know, I don't know, quitting in protest, sabotage? I'm sure you're like, it it has to run the gamut. And and I wonder if there's any sort of general purpose lessons that you would, you know, share with with just business leaders in general regardless of whether they're entering into a structure like this. General purpose takeaways for like AI transformation.
00:39:43 · Marc Bhargava
[00:39:43] S1: Yeah, absolutely. I think there is, you know, at first a bit of apprehension around any new tool or technology. One of the really distinctive things of what we're doing is they're saying we're coming in here to AI enable the workforce. So we are not coming in here to replace the workforce but to AI enable it and to take on more customers and have more revenue and have a growth story. And I think it comes across as pretty authentic because we're generally not using debt especially in these early transactions. So we don't have these interest payments to make which means we have to cost cut, which means we have to look for savings. We're coming in and adding costs in the form of software subscriptions, engineers, but we are placing this bet that an AI enabled workforce can do a lot more work and have a lot more revenue. So to the staff or to the folks working there, is a challenge of you know we have to get people up to speed on this AI technology, we're going give you these tools but then of course you also have to go out and get more clients and get more revenue. Generally though, we're saving them from the most mundane or boring parts of their day or the most repetitive parts of their day. So we are freeing them up to do generally more interesting, more demanding and more human to human kind of interactive tasks. And so once the process is going and you know as you're rolling up more and more companies, it really becomes much easier because we many times once you've done a first acquisition, we have those teams especially the leadership team but even people at those companies talk to the second team and help win them over for the second acquisition. And so that I would say is a really strategic part of it as well. Once you've done a few of these, you have the stories and the case studies to tell and I think that really assuages a lot of the concerns like you mentioned like, oh, are people gonna get fired or is all of this gonna get automated? It's like, no, we've done two or three of these, here's exactly what happens. We give you the software and tools to automate your mundane tasks, you get to do harder workflows, you know, it ends up being better for the company, your shares are worth more. And so it's very compelling I think compared to private equity alternative especially.
00:41:36 · Nathan Labenz
[00:41:36] S0: Are there any like rules of thumb or like an example from the ambiance Cleveland Clinic partnership was Cleveland Clinic required that every doctor in their system, I think, use the AI scribe once. And I thought that was really interesting because obviously these AI tools, you know, have a... Have an error rate. So that sort of projected confidence, it was like a minimal ask. It says a lot about like how reliable, you know, you might need your product to be if you're gonna make those kinds of relatively minimalist demands on your team in terms of, like, trying it out. In that case, they... From that one interaction, they report that they now have 75% of their 4,000 doctors using it all the time across 60 specialties. So presumably, that's like a product excellence translated to like this is amazing sort of story. I assume one of the advantages of being like more tightly integrated is like you don't necessarily have to get to that level before you can, like, kind of push the adoption. But are there any, yeah, like, mental models, rules of thumb, frameworks for how to think about how far the product has to be along versus how much to push people to use it even if it's imperfect? What what are the sort of managerial takeaways that you've seen?
00:42:57 · Marc Bhargava
[00:42:57] S1: I think a big learning for us is to find the assets and the companies that have a pull factor. So if you're purely underwriting on a pre... PE style, maybe you're putting in a new management team, maybe you're cutting costs, you're doing all these things. You... That's kinda what you're looking for in the asset. For us, it's really important. Does the company... We've now incubated a company. We're going to buy one of our clients. Does this client really want the technology transformation? Have they tried different products? Have they tried hiring people? Have they really pushed through at the leadership level that they want technology and AI? And we will pay a premium or our companies pay a premium in the acquisition to have the right partner and the right partner mindset. And so that's one thing that's extremely important is from the top, from the leadership position, this is something they've been trying. And then there are all kinds of techniques once the acquisition has been made to of course have it not just leaderships buy in but the entire company. And that honestly is not changing up the workflows too much, keeping it similar. So like in HOA management, you know you can email saying, hey create a version of the HOA board deck for me and it will create it and send it to you via email. So we're not trying to onboard you to some new software tool. In that case, know, you're using your existing text or email workflows. So earlier question you had is like do we totally reimagine it? Not really. We wanna keep a lot of the same workflows but we've already identified behind the scenes a huge part of it, 30% plus can be automated. And so we're trying to keep folks' workflows very similar but use AI with sometimes them not even knowing that you know what's the large language model doing behind the scenes and how is it writing this copy and all that stuff. It feels like they're just interacting with an agent that is able to send them a board deck that they need for tomorrow's meeting, etcetera. So we're generally mapping over those same workflows and so it's not a big change many times for the workforce to actually learn and adopt these toolings because our companies are building the tooling to be very similar to what