Revolutionizing Reporting: Steve Wasick Unveils infoSentience's AI-Powered Breakthrough

0:00What can 160 years of experience teach you about the future? When it comes to protecting what matters, Pacific Life provides life insurance, retirement income, and employee benefits for people and businesses building a more confident tomorrow. Strategies rooted in strength and backed by experience. Ask a financial professional how Pacific Life can help you today. Pacific Life Insurance Company, Omaha, Nebraska, and in New York. Pacific Life and Annuity, Phoenix, Arizona. The wait is over. Dive into Audible's most anticipated collection. The Best of 2025, featuring top audiobooks, podcasts, and originals across all genres.

0:41Our editors have carefully curated this year's must-listens, from brilliant hidden gems to the buzziest new releases. Every title in this collection has earned its spot. This is your go-to for the absolute best in 2025 audio entertainment. Whether you love thrillers, romance or nonfiction, your next favorite listen awaits. Discover why there's more to imagine when you listen at audible.com slash best of the year. Welcome to the AI Chat Podcast. I'm your host, Jaden Schaefer. Today on the podcast, we have the pleasure of being joined with Steve Wasak. Steve is an incredible entrepreneur, founder, and he's going to be telling us a little bit about what he is currently building.

1:23Thanks so much for coming on the show today with us, Steve. Hey, Jaden, thanks so much for having me. Would you mind telling everyone a little bit about your background and about your company? Yeah, I think I have a little bit of a strange background for a tech founder. I was actually in my third year of law school at Northwestern when I had this idea for sort of like fantasy sports newsletter. Okay. Which I thought we could do sort of using automated technology. And I started working on it like in my free time. And I quickly sort of realized, hey, like this basic technology could be used for all sorts of applications, finance or medicine.

2:01Anytime you have a big data set that's complicated and people will sort of want a synthesis, like a quick update of what's going on, we could use this technology. So instead of, you know, becoming a lawyer, I raised a little bit of money and started this company called Imposentience back in 2012. Okay. And yeah, so that's my background. And I just took computer programming in high school and just kind of kept up on it a little bit. But it's more what I'm doing or what the company's been doing is pretty brand new. So there's not like, oh, I missed this class in comp sci when I was going to school there.

2:38It's like there is no class on it. You're all just kind of making up stuff. And it's all based off of logic and narratives and things like that that actually my background, which is I had like an English degree undergrad and then a lot. It's like, Hey, it's like you, I think it's the combination of like English stuff, logical thinking, and then like the actual computer technology stuff. It's like all three of those. So it's like, I had a background in two. And so I don't feel like I necessarily missed out. Like if I had, you know, just done comp sci or something, I feel like I would have missed on some difficult things like that you have to deal with, like in terms of like just having good writing, you know what I mean?

3:17So I think it's a little bit of a different background, but for what I'm doing, what the company's been doing, it's kind of just as good as anything else. Yeah, for sure. Well, that's really interesting. So, you know, coming to the present day now, you're running InfoSentience. Would you mind telling everyone a little bit about the core problem that you're solving for InfoSentience, how AI plays into that and kind of what you're doing? Yeah, I think the core problem is just data overwhelm, right? So we've had all these tools that have grown that have allowed people to store so much more data than we ever had before.

3:50But the tools to analyze that data haven't necessarily kept pace. So there's tons of companies, organizations out there that have a tremendous amount of data. Sometimes it's in different silos. Sometimes it's in different formats. But they'd love to get analysis of that, especially analysis at scale. So as an example, you know, our first product was in the fantasy sports space, which is something that we still do for CBS Sports. And for people who are unfamiliar with fantasy sports, it kind of works like a stock portfolio. You sort of pick a group of players that you think are going to be better than some other group of players.

4:22But every single person's portfolio is unique, right, as millions of players. So you can't cover that with human beings, right? Right. It's impossible. So that's a situation where our software comes into play, looks at all the data, tries to figure out what's most interesting, and then writes up a report that reads just like a human being wrote it. That's compelling. That offers insight. That's different from week to week. Just like you would get if you had a human being just covering your fantasy sports league. So that's what was our first product. But all of our other products are similar, right?

4:58Where it's just like people have a lot of data. Like we have, we, we work with the Chicago Mercantile Exchange to analyze all the commodity data. We work with medical companies, retail companies, right? Where it's like, you've got a hundred products that are out, right? And they're all moving every single day, right? And they're, they could all be distilled down to different regions or different product groups or like different, there's different metrics, right? It's not just sales, it's the returns, it's discounts and all sorts of things, right? So you have all these metrics of all these products moving every single day.

5:32And the question is, how do you get a handle on that? And right now, the answer is typically, oh, okay, at the end of the month, we kind of talk about it. Or maybe we meet with certain teams once a week. Our system, it's like, all right, do you want a daily report? Do you want it updated every hour? Do you want it focused on a particular topic, whatever breakdown you want? Like, I want a report on the U.S., West region, these types of products over the last three days. Give me that report, like distilled. right like the most important thing in that information because you can kind of do some of that with pivot tables yeah but that is it's not going to give you the context right it's not going to say like oh here's exactly what was driving it or here's how it fits into you know the last month or the last year or like what changed in this three days that's like you are making this difference so that's all the type of stuff that a good human analyst can do and we can impart all that intelligence into our software so that it basically can work the same way.

6:28Very cool. So how does, you know, AI play a role in what you guys are doing today or what you guys plan on doing in the future? Honestly, we took a different tack than I think most companies did when we started. And certainly most companies now, it's like everything is like LLMs, right? So we don't use LLMs. Our system is non-probabilistic. We use what we call conceptual automata which is basically taking all the events that a human analyst would look at and breaking them down into their subcomponents and then putting the intelligence as like the modeling sort of intelligence on those subcomponents and then they have sort of the ability to self-assemble into a larger narrator so i'll give you an example you know first from sports for instance like you have the concept of a team right and you can put a bunch of intelligence and sort of modeling and hierarchy stuff information on that and then you have the concept of a team winning a game right that's like an event and then you have a concept of a streak right which could be for sports or it could be anything else it's really just like an event taking place again and again over time well you can put all those together and now you have like a team on a winning streak right but like when you assemble those together that grouping now has all the can grab all the like intelligence that's on just like the concept of a streak.

7:47Like for instance, what, what do I, uh, to avoid, let's say repetition, right? You know what I mean? It's like, okay, well, some different team is on a streak or maybe they're on a losing streak. Right. So that's a different concept, but there's still an element of repetition to that. It's like, is that really the best story to follow up? Or you have a situation where if you say, Hey, this team's on a losing streak. And then the next story is the team team is on a winning streak. You're going to write that up differently. It's like, you have to say like, however, or something, you might have to have a transition because these are substantively the same story, right?

8:17And that's just a very simplified example. But the idea is that human beings are really, really good at just adding a bunch of concepts together and seeing how they all interact. But that's something that traditionally computers have not been very good at. And the problem is because the number of interactions starts to get really crazy as you add more and more things together. And so the solution that we found to get around that again is to kind of put the intelligence on the low level so that all these things can, you know, you can kind of put it in this big soup of concepts and they can all kind of like communicate with each other, kind of like what neurons would do in a human brain.

8:55And so that has been sort of what we've used traditionally now. These LLM tools are obviously like pretty amazing. So we're certainly exploring how we can integrate some of what we're doing with LLMs. but obviously they currently right now kind of struggle with data, struggle with operational operations. I mean, they can do them, but it's not that their forte, right? And some of the things that we're doing, which are really sort of complicated analytics, particularly around, let's say, proprietary data, where there's not like a corpus of text that's talking about how this is being analyzed. That's something where LLMs, at least as of right now, are not particularly good at solving problems, right?

9:40And they also have like solutions, which you were about to run it, debugability. I mean, there's all sorts of things where if you're doing like what we're doing with the CME, where we're taking in 50 gigabytes of data and instantly sort of giving analysis to it that's like on every single topic. Yeah. That's just not something that LLMs are particularly good at right now. Yeah, and this is so interesting because, you know, of course, I talked to a lot of people integrating AI into their companies and LLMs are kind of the hot thing everyone's talking about and doing right now. This is really interesting because, you know, it would appear that you're doing this in a little bit of a unique approach to other people.

10:16I wonder, is this because when you kind of started, right, you said you got this company kicked off back in 2012. This is because when you started, I mean, obviously LLMs weren't kind of this big thing. It wasn't this possibility. So do you see that shifting a lot for you in the future? Or do you think, you know, this kind of current approach that you have is one that's going to be the winner long term, right? Other companies might be looking at this, they're debating, like, should we come up with something more like what you've done? Or should we come up with something that's more LLM focused?

10:43Where do you kind of see this going in the future? I don't know. I mean, I think emerging is probably the right way to do it, like between the two technologies. That's what we're trying to do, because I think that there are pretty clear strengths and weaknesses of both of them i mean in the long term who knows you know what i mean like it's like that the llms are so some of the things that they do are so extraordinary and so indicative of like really high level intelligence and i feel like so much money is getting poured into them so where they yeah well some means there are there's going to be improvements inevitably eventually so yeah so i mean i think you know maybe 10 years from now it's all the llms right like but down the efficiency and the but 10 years is a long time you know what i mean so it's like And the other thing is the way that we really see our technology is if we're partnering with a client or potentially maybe if we were ever to get bought out or something, it's like we do view our technology as a really big accelerant, right?

11:39So it's like we can do some things right now that LLMs can't do. So even eventually if like, let's say we got bought out and then six years from now or something like our technology isn't being used anymore. It's like the fact that we were able to provide that service, right, for these clients, right? Because there's so many things beyond the LLM. It's just like, right, pulling down data. Are you trustworthy? Like, do we, you know, like, do we think that you're a good person to work with and, and, and, and you've sort of like figured out like the, the things that are particular to our use case that are really valuable and how to deliver it.

12:14And all these other things that like are not necessarily really going to be working with LLMs. Like just the pure data analysis part of it is only a subcomponent. Yeah. And so we do think that regardless of whether our technology is like going to be used in the long term, it's like it can definitely do things that LLMs can't do right now. And so we're trying to just build up that client list and build up those relationships, right, with different companies in a bunch of different verticals to say like, look, at least right now, this is the technology. And as the LLMs get better, you know, we intend on incorporating that as much as we can.

12:49Right. And because, yeah, so exactly how they're merged, I don't know. We're still like early on in that. Yeah. We do think that like using our technology and LLMs, we definitely think that there's a space for it. Yeah. And I also think that, you know, what you're doing is quite good because obviously there's a lot of people right now that are struggling with the whole data aspect of the LLMs. You know, there's different workarounds or there's plugins where people are trying to integrate things to help their LLMs do better with data or math or other things. but like it's not you know the core yeah it's the uh chat gpt for example um and so you know having this built out your own way is uh is a really good move what would you say is one of the hardest challenges that you guys have kind of overcome or are you know grapple with in in this whole business and industry i think just the the specificity of data you know i mean this is something that companies deal with all the time where it's even internally right where it's hard to even within an organization have the same types of data the same consistency of um of data right where it's like so this organization does this or this organization does that or then also like debbie in accounting has the her spreadsheet where she does this one cup subcomponent and so for our system which is all based off of structured data ultimately right we have to ingest it and um do so in a way that's that's pretty consistent and so a lot of times what happens too because Because our system is really good at essentially finding unique pieces of information and essentially we might have 200 different modules of analytics that are being run.

14:31It's like they're all run at the same time. And so what ends up happening a lot of times when we work with a company is that we'll find the errors in their data or the missing parts of their data, right? Which they weren't even aware of because he was even looking at it, right? Like now that we're doing a deep dive, you know, it's like, oh, do you know that this actually, you know, nobody puts a date, you know, for this data here. So you can't actually put it into like a time series. A lot of things like that are really difficult. So a lot of times we'll talk with a company and we'll come up with like a plan.

15:01And then a lot of times once we actually get into the data, it's a lot more complicated. and and then some some of those things you can fix by having better data but other things it requires us to really put in a lot of effort to have the near the like the analytics side of the engine being much more flexible right to take into account like oh okay sure there's these 10 things but some of them are going to be weighted really differently or so you know it's like all these things that are again for human beings it's not that hard but like when you get into a piece of software, having this type of human, human-less type flexibility, being like, oh yeah, that's the way it is 90 % of the time, but then this other 10 % of the time, you kind of have to flip how you're viewing, you know, the 10 products or something, or this is like, this is Christmas time, so actually like these numbers are very different during this time only, or we do this, this is our year end something, something, whatever it is.

15:56A lot of times there's these little quirks in the data that if you're going to have a really solid narrative, that people can really use, you're going to have to take all these little quirks into account. Yeah, that does sound like a challenge. Tell me about what your thinking is from kind of a broad, a high level, right? There's a lot of people, obviously, this is something that for, I would say for the mainstream is a kind of a new concern, right? As of ChaiJupti, as of this year. But obviously, the issue has been around for 10 years. There's a lot of people that are very concerned, like you know ai for example is going to take the job of a data analyst um you know some sort of software technology uh that you're building is going to you know take someone's job what what is your what are your what's your thinking what's your response to to people that might have some sort of concerns in this area i mean definitely the software we have right now is not going to take over everybody's job um again like an llm that can really just analyze any data set just sort of spontaneously and no matter the size and come up with narratives and visualizations and everything i mean that's at some point yeah that maybe probably does cause a lot of problems i mean i would just say like what we do right now for instance with cbs where we actually write articles on the nba the nfl's you know all these different live sports leagues right huh um but they haven't really taken over any human jobs it's more just like they're still writing like the preview for the super bowl for instance like it's not being done by our system right like but you know some random you know mid-atlantic conference basketball game on tuesday it's like well they didn't have a reporter that was writing a bunch of content about that game right right so it's like we're sort of filling in the gaps i guess to a little extent like they've relied on our system to cover games that they might have had a human being cover a little bit.

17:50But I think the reporters in general are very happy to be doing something that's not like as rote repetitions. Right. They're then doing stuff about like some interesting off-season trade or some sort of interesting statistical pattern or some sort of thing with the coaching. You know, something that's more unique than just like, here's the 500th preview of a basketball game that I've had to write this month. So at least right now where we sit in the value chain, like we're not you know we're not necessarily taking over from what analysts do and even even in like the business analytics side of things like if it's like a quarterly report like we're not going to do as good of a job as somebody who that's their whole thing is they just look at your company's data and they know everything that's going on right like so that's not what we try to compete with it's more just like all right but if you want a daily report that's going out to all of your sales people and you have 300 sales people like that's just not going to happen with human analyst, right?

18:46But they might really have a lot of benefit from seeing like the different insights, right, that we can pull through and deliver something specific for each person, right? And so that's really where we see the value proposition of our technology. And I don't think at least at this point that that's something that's actually displacing existing human data. Yes. Okay. That's interesting. So, you know, talking about, you know, you have kind of like some automated articles and stuff generated talking about different basketball games or different sports leagues or different areas that, you know, there's probably not enough firepower bandwidth for a certain organization to cover, right?

19:22How do you kind of handle the challenge of keeping this AI-generated content, for example, engaging and not sounding too robotic? Yes. I mean, that's a great question. And I think we do a much better job than our competition in this area because we don't use like templates, right? Or what I call like the Mad Libs approach, which is just like, hey, here's a narrative. We're going to fill it in with like these basic details. And it becomes pretty obvious when you're reading that, that that's what you're reading. Like, even though there's no like grammatical errors and it's like, it kind of reads fine, but it's just like, there's an aspect to it that's so rote and basic that you pick up on it really quick and it's, and it feels boring.

20:02And using our system of conceptual automata, where we have all these little pieces, basically what we do is we tell the system, Hey, figure out every single interesting thing that happened. like in like in a sports game okay but there's might be like hundreds of things right and you kind of like rank order them and then give them the ability to sort of sort themselves out so so that you can get the unique things that happen because in a template the problem is you had to put in the template all the the things that are bound to happen right somebody's gonna win somebody's gonna lose there was a best player and here's who they play next week right like it's all this basic stuff but you're not going to put in like oh this is like a wild comeback and here's what happened and you know all these different things are like man like this you know this team had one four straight in a row against this team but now they came back and this is the first time well all these different like all the unique things are the things that are actually interesting yeah and so that's where if you have flexibility you can surface those things right in such a way that because that's what people respond to right and so they they they can um you know they can really pick up on that i think very quickly that's that's super interesting and that's super smart again uh just coming up with the most interesting things and doing the ranking system and having that in there um yeah people definitely pick up nobody likes a template no one likes it to sound too robotic um and so like people will still really enjoy consuming that so that that's very interesting and they'll get great insights out of it as well yeah what what industries or sectors do you see benefiting the most from your technology that you're currently building, whether that's one you're currently working on or an area you see in the future?

21:37I don't know. I don't know. Hopefully somebody can tell me. I mean, we're in a lot of verticals right now and it's really hard to say. It's really hard to say because it's hard for us. This is another challenge that we face as a company is that we can look at industries or companies from outside and kind of have a guess of what data they might have, but the reporting needs are definitely different. different, right? Like, it's like, do you really need a daily report? Do you really need a report for all your salespeople every single day or every single week? That's where it's really difficult to know, you know, exactly when, when we can really have a value proposition that helps people is, is tricky to know from the outside.

22:16So our, our approach is kind of just scattershot. I'm just like, all right, well, let's just talk to anybody. Um, and we have, you know, we have multiple companies in the finance vertical. We, we work, we have, you know, a medical client, we have a retail client, we have a marketing client, we have, um, you know, journalists like CBS and some other folks, actually multiple folks in that space. So we just try to kind of talk to, talk to people and, uh, because we don't, we don't know the answer to that question on it. Yeah. Like what would be like the best? I don't know. Yeah. Well, I mean, I think it's applicable in a lot of different areas.

22:52But talk to me about a couple different case studies or success stories. I mean, even just for people listening to this to kind of conceptualize some of the areas that you're making some of this innovation. So like in finance, for example, what are people using this for in finance? Well, the big thing that we're using this for is with the Chicago Mercantile Exchange, which we were doing sort of a demo product with them for a while, summarizing some different commodities in the agricultural space. Yeah. We're going to tell it changes that the largest commodities trading firm in the world. So they do options and futures contracts around, you know, like soybeans and gold and interest rates, everything like you can kind of just like bet on it essentially.

23:32Okay. And what we're going to be doing for them though, is something I think really cool, which is basically taking all their commodity data and turning it into an always live, always correct web page. Oh, that's, uh, that's basically looks like a sort of Yahoo finance or Bloomberg, uh, uh, web page, but everything is automatic. So you see is like always up to date. So it's like, you're seeing headlines and pictures and videos and narratives and everything on every single topic within the CME. But it's, it's always sort of now, right. Or at least radical, you're live. You're always seeing everything there that's now.

24:12And if something, for instance, if a commodity goes way up and then it goes back down and now it's just back to even, right? It's like, maybe that story disappears off the front page, right? Because that's not the story, right? Which is not what happens on traditional normal. Yeah. It's like you have a brawl and like something's there and maybe it's hard to kind of say like, oh, this is what's happening with this stock because by the time somebody reads it, it might be something different, right? So that is actually very innovative. and i absolutely yeah because i definitely hate clicking on like a an article that was like this incredible thing happened and then you go check out the stock or whatever and it's like back to back and not yeah i that let's see now it's a different story though because now it's like hey it was up and now it gave back all its gains right like so it might still be on the front page but it's like the up-to-date version of that and so i think that it's and because we're providing so much content one of the things that we're gonna be able to do is what we call like this dive deep dive technology, which is like anything that you're looking at, you can kind of be like, oh, that's interesting.

25:13Give me more information just on that. And they'll like take you to like a new article that's just like focused on that particular topic. So I think that that'll be pretty cool. Yeah. You know, the thing we've done with CBS, which is our longest client, you know, the fantasy sports, people really love it because they get this customized insight and And they get like an email every week. And those are like the highest open rate emails that CBS has of an email that they send down. Because people are like, it's just different because they're getting something that's just for that. Right? Like nobody else is reading this.

25:46It's just this whole article. It's like 300 words with pictures and highlights. It's possible a single person could service those customers with something custom. So, yeah, that is really amazing. Yeah. So that's, I mean, I think that the power of customization and again, this power of getting away from the robotic sounding template stuff. It's like when something, when you can write something that really speaks to somebody and really gives them the information that they want to hear, that's unique to them. Right. It's like, it's really powerful. And so like a lot of our, of our customers have, have experienced that and, and it's been, it's been fun to build.

26:22That's really interesting. What about, I think you mentioned like healthcare. How are you seeing this in like the healthcare space? Our healthcare stuff is pretty early on. I mean, the only client that we currently have is IU Health, which is like the largest hospital per puric in Indiana. But basically what we do for them is automated doctor bios. So they have biographies for their doctors, but there's thousands of them. They're constantly coming and going. The data behind their bios is changing, right? Like they might change locations or specialties or all sorts of things. and so what our system does is it takes all their information like their education you know whether they speak different languages and also one of the really cool things it does is like actually synthesizes patient reviews right so we like pull out some of like the key things and also include some like curated examples of of new patient reviews which is good for seo right because the articles are kind of changing and updating and so it's just been a way for them to uh really save a lot of time and actually a lot of times they just didn't even have a bio for doctors which is not good and people like to to read something about like a summary of who this doctor is and the medical history and if they have like publications like it can include that so it's pretty flexible um but we haven't i mean we'd love to be able to do more stuff in in the medical space but it's tricky because of all the patient protections and obviously the the um it's very important to be correct, right?

27:48So if you're doing like actual like diagnosis stuff, so it's a little bit harder to break into. And we've had some good conversations, but we haven't really done as much as we'd like to do in that space. But hopefully in the future, we can do more. So yeah, that is interesting. And definitely, there's some nuances and challenges in that industry, for sure. I think a lot of people are grappling with those right now. So nothing unique to you. So I saw, I believe it's on your website there's a spot the bot challenge yeah yeah yeah um what kind of inspired that what's been the feedback from users on that i i mean i'll everyone that's listening what that is yeah basically we just took a couple of of data sets that and and looked at okay here's what a human being wrote up on this particular data set and here's what our system did right using actually a finance example and like a i think it was like a new york giants game or something like that where it's like here's the ap report and here's our what we wrote for cbs and um yeah i think you know we just started doing that at like trade shows kind of we were just we just actually printed out as like kind of a game just to get okay come over and then it was like oh we should put this up on the website and it's fun because people always a lot of people comment on they're like i failed or something you know like they said i didn't get it right um and so it's it's nice because it really does bring home this idea of like hey this is it's it feels real you know what i mean like and again this it comes from picking up on some pretty subtle things in terms of like the structure of the narrative and the sort of uniqueness of what it's talking about and it allows it to kind of uh hopefully in a quick a very quick easy way get people to understand the the difference between what we do and maybe some more simplistic solutions out there yeah for sure that's awesome um so how many people are currently on your team uh just five we're small okay that's awesome a lean team and you've been going for you know the company's over 10 years old now right so that's amazing um what has been yeah how has that been in the process of kind of growing and scaling the team from early on were you the original like developer on this did you bring people on how how has that been yeah i was the original developer and then i brought in just one other person and it was just that for for quite a while we've had a really sort of weird journey because we started off we got cbs as a client right away which was sort of a miracle i don't know how the heck that happened looking back on it but it happened and about like a year in we had launched and i think we had proven out our technology to be really good and and i would say industry leading in certain ways but i had this idea for like a better version which i thought would take like six months and it took eight years and that is classic uh development problem yes and the whole time i thought i was like almost done so it wasn't just like oh if i knew in an advance it's gonna take eight years like i could kind of plan ahead but like i always thought i was like almost there you know and then i would have to like restart over from scratch and so So we didn't try that hard to expand the company.

30:59We got some clients haphazardly, but we didn't want to make a big push because it was like, well, we've got this new technology. So we want to build the new clients on the new tech stack, which is almost done. So let's just wait three months, right? So it was just like that for a long time. But thankfully, about a year and a half ago or so, we really got the tech to do everything that we ever wanted it to do which was great um eighth times a charm like almost it literally was like almost every single year we started over from scratch oh my god um but we got it working now and um so yeah so then we've kind of shifted into trying to be like a real company as opposed to just like a research project disguised as a company which is what we were for a long time and uh and so now we yeah we tried to expand but that's kind of like why you know we we haven't grown as much but honestly i mean like with the team that we have right now i mean we could probably double our revenue before we'd hire another person i mean like because we're still like even with the team we have right now we're still doing a lot of kind of porting over some of our own like porting over some of our old products onto the new system things like that like we have a lot of a pretty good amount of slack in terms of like even with the five people because the tech is really good like it allows us to build really fast and and we just have all of our back-end systems in place like the ability to like do this the scale you know spool of instances on the cloud like debug it make sure that we're delivering what we were supposed to be delivering like all those procedures are pretty central so it's like we're we're in a good spot to to grow even more that's really exciting um talk to me about so i know at the very beginning you said you you wanted to get the company kicked off so you went and got some investors was that your only round of funding have you done multiple rounds of funding as you've been growing no that's it like it was just like a small friends and family round and that got us where we need to be because we got cbs like right off the bat i mean i didn't know what i was i still don't know what i'm doing but like i didn't know how funding works like i mean i gave away like an absurd amount of the company like for like not a lot of money i didn't know yeah and uh so so thankfully you know we've been able to just sort of uh bootstrap since then and we're profitable now we're not looking to i mean we couldn't like i said it's like we could double our business with our current with what we have so it's it's we don't need to uh raise any money for anything that's super cool that's that's amazing and honestly a really good place uh talking to a lot of people those lean teams are sometimes the it's the most powerful way to do And how has it been for you?

33:38So are you like, would you call yourself like the CTO as well? Like, are you doing the main development? So how has it been kind of being CTO, CEO, like, like kind of being wearing all the shoes in this? I mean, you know, as probably as you'd expect it to be, it's a little crazy. And obviously you feel the limitations, like your own limitations, right? Like I, you know, I like to think I'm okay at doing, wearing a few different hats, but like you're never going to be as good as somebody who like, that's what they do, right? Is like sales or they do marketing or anything else. So, um, it's definitely been a learning process, right?

34:16Like, and you can, the scary thing is like, I look back on what I knew, let's say five years ago. And I'm just like, man, like I didn't know anything like, and I've been an entrepreneur pretty much my whole life. So this is like, I've pretty much just been an entrepreneur for like 20 years straight. And like even five years ago, it's like 15 years into the process. Like I didn't know anything. Right. And now it's like, well, I feel like I know a lot more, but it's also scary because it's like, well, yeah, but like maybe five years from now, you'll be looking back and be like, my goodness, like how stupid can you be about X, Y, Z?

34:47So it's like, it's cool to feel like you're growing, but then it's also nerve wracking to be like, well, I'm sure I'm just as ignorant about like lots of different things right now. and and obviously the more hats you wear like the more things you're going to be a rookie right doing so i don't know we're still here we're in business we're profitable we're growing but like every day is an adventure yeah it's a little different kind of in that line of thinking right um just in crazy incredible learning curves and like even looking back five years you see how much you've come what's a piece of advice that you could give to um you know founders today that are working on an ai startup that are working on a on a company uh what's a piece of advice you might be able to give them i mean it's pretty probably pretty generic but just like customer focused like start with the problem as much as you can i mean i i'm like i like technology so like i definitely get caught up in like oh wouldn't this be cool and it's like yeah it's cool but like does it actually is it going to drive value is it going to be easy to scale is it going to be you know all these different things right so yeah i i think especially if you're a first-time entrepreneur like start with a really small problem and especially something where if it could be like niche yeah you know what i mean like i remember this one guy who i knew he was working as like a just like just on spec for some project and he built this like uh some sort of system for like people who run dog kennels right okay and it was just like built it up for this guy who's who'd been an entrepreneur for and like he you know he's just hired to do it i think he did in like six months and then this company was worth like it was getting offers like three or four million dollars you know like to do doc like because just like yeah nobody else was doing it and so like they were able to just take over the market because they had this this specific expertise right in this one this person who started it was like i know for a fact that people need it they probably did like some pre-sales or a lot of talking to folks but i think that um that's definitely that that's that's the one of the big mistakes i made i didn't even have client when i started this company you know i just was like oh this is cool tech you know and like like i said it was sort of a miracle that that we got not only a client but like cbs is a huge client right and so you know don't rely on that miracle right like you've got to try to um again yeah figure out exactly the problem you're going to solve which i know this is all pretty trite advice but i it's it's common advice because it's so important i would say well yeah no i mean I do think that is really spot on.