Tj_Vantoll:
Hey, everybody, and welcome to another episode of React Round up. I am your host today,
Jack_Herrington:
Yes,
Tj_Vantoll:
T. J. Van Toll, and with me on the panel, I have Page needing house
Paige_Niedringhaus:
Hey, everyone,
Tj_Vantoll:
And Jack Harrington,
Jack_Herrington:
Hello, everybody,
Tj_Vantoll:
and our special guest today is Shawn Austin, Shawn. welcome to the show.
Sean_Austin:
Great to be here.
Tj_Vantoll:
Hey, so maybe you can start by telling us a little bit about who you are. what you do. Why
Jack_Herrington:
Bis,
Tj_Vantoll:
you're famous. What we're talking about, All those good sorts of things.
Sean_Austin:
Sure, famous is is the one I use only only a little bit now, but hopefully hopefully more so. After this
Jack_Herrington:
Yeah,
Tj_Vantoll:
Uh,
Sean_Austin:
podcast We.
Tj_Vantoll:
uh,
Sean_Austin:
so, I mean, my career has been all and actually mostly product technology right very much around soft where trying to go up the ladder for impact and actually just make better products in the end of the day, so started not to go too deep in the weeds. But just trying a bit about me is as actually is where I started to the trends of, as I think, Lucky for my my year of birth, where I came right, An Android got launched and there was this huge like Sunamiwave. Now I find myself seeing it with a I, and actually working on a voice. a company that's very much around audio, so kind of the N to the language side, like the End of the Yang, and a lot of experiences from Spotified Clara in Sweden, and just after products and trying to build startups,
Tj_Vantoll:
It's very cool. We had a sort of prediction show at the end of the year and we predicted our theme of the year was a I, and I believe it might have been jacked to that, Said like, if you're
Jack_Herrington:
Starting
Tj_Vantoll:
starting a start up right now, A
Jack_Herrington:
Is
Tj_Vantoll:
is like the place to be right, so I'm assuming you were listening to our podcast
Paige_Niedringhaus:
Uh,
Tj_Vantoll:
and
Paige_Niedringhaus:
uh,
Tj_Vantoll:
following following our advice.
Sean_Austin:
Sounds like you nailed it
Jack_Herrington:
All the F. T.
Sean_Austin:
perfectly.
Jack_Herrington:
C. Just put out a warning. Yeah,
Tj_Vantoll:
Yeah,
Jack_Herrington:
did.
Tj_Vantoll:
so maybe you can start by just telling us you what got you personally interested in this space right like, and maybe then lead that into Like what it is you're doing Now? Like what? what company your working on? what sort of product you're offering? That sort of thing.
Sean_Austin:
Yeah, so I feel like you definitely nailed it, though, with that prediction, so I can't wait for what's coming this year. I need to make sure I m getting the inside information on it, but we or I actually
Jack_Herrington:
Yeah, yeah,
Tj_Vantoll:
hm,
Paige_Niedringhaus:
Yeah,
Sean_Austin:
mean my co founder, and I want to say we. we. We thought about a. Like. Basically the terms of big data. So just how can you drive intelligence? Since A has been a term for for quite some time. He
Jack_Herrington:
Hoops.
Sean_Austin:
There we go. so I think I think I may have just
Paige_Niedringhaus:
Hm
Sean_Austin:
started for a second, virtually stuttered. But we we thought about it. Like from big data intelligence right, What's what's possible with all this data that's available In the fact that now compute. so, thinking about trends, thinking about it from that layer, and I am totally biased towards the audio because now ten years ago was in the first acquisation was spotify, So really like my twenties were around, Big data with Clarina. Audio was spotifyed for about four years. Um, and for me it's like, so I'm just sort of was exposed to it. It quite a bit. but I've heard the term a lot. I actually now feel like it has a whole new whole new meeting. Like language now is front center fra. But that that trend was definitely one that's been around for a bit. It was growing. It was accelerating. the idea that you can like mine information to get intelligence from it, or higher level insight was available, and then back to the audio lean. So we try to do. Try to do something Ay that's not been done. So will we focus on what Helios? So Helios Life Enterprises is the start up which has been around for a couple of years. We focus on the tonality, so the sound of voice, not the language. So when you think about this conversation
Jack_Herrington:
Huh,
Sean_Austin:
asking me a question about the future, I may or may have not have a lot of confidence in my answer Right, like Cos, are always trying to be confident or train to be in the rewards as confident as possible. Like everything is rosy right. So we sort of had this this high moment. We said Well, If you could get to the tone of voice, it's not so much. Is it like a polygraph? But it's more. Can you supplement your understanding of that C. O? Who's saying we're going to ship five hundred fifty thousand cars at a China next year, Right like there's a layer of understanding we get as human beings that we all know From sarcasm to you know, well, confidence or everything in between. That tone of voice is a big chunk of the message. so we got you know. that's how it started. We've luckily signed up some really large hedge funds. We believe all of finance really needs this supplemental information.
Jack_Herrington:
Fascinating. So how do you prove that you're actually finding out any additional information?
Paige_Niedringhaus:
That
Sean_Austin:
Yeah,
Paige_Niedringhaus:
you're
Sean_Austin:
without
Paige_Niedringhaus:
accurate.
Sean_Austin:
getting.
Jack_Herrington:
Yeah,
Sean_Austin:
Yeah, right
Tj_Vantoll:
M.
Sean_Austin:
right. So a lot of wishful thinking early on because it's a
Jack_Herrington:
Uh,
Sean_Austin:
science endeavor.
Jack_Herrington:
uh,
Sean_Austin:
But the big piece is customers coming back and actually purchasing this from us over and over again. You know, like renewals. So what we've done is focused on the customer side. I think I jumped again. I'm not sure what's happening with that audio
Jack_Herrington:
Yeah,
Sean_Austin:
page.
Tj_Vantoll:
Yeah, you might need
Jack_Herrington:
yeah,
Tj_Vantoll:
to.
Jack_Herrington:
you need
Tj_Vantoll:
You
Jack_Herrington:
to
Tj_Vantoll:
might need to just
Jack_Herrington:
start from the beginning.
Tj_Vantoll:
well edit this. You might just start with that. that answer from the top.
Sean_Austin:
Yep, which is weird, because let me see if I can take something of life. So that Yeah, So how do we know it's accute? Starting from there, we have tried so many different ways of simulating portfolios, So when you think about accuracy in the space, it can you take a portfolio A way. That's basically how do you trade these securities, these companies and make it better, So there's a lot of lot of math and a lot of testing that can go into that, and actually probably the best indicate Is a customer coming back. You know, very large customer coming back and buying this again, right so they won't tell you how they do their portfolio, But what they will tell you is we want to buy that again, so luckily we've had very high renewal rates. We have done a ton of internal testing, and now for the first time we actually have handed this of to like research universities and some other very capable, sort of like third parties. You can think about it to clients, but third parties who can, who can check us right. So that
Paige_Niedringhaus:
Hm,
Sean_Austin:
is Big piece of this year. So we've you now seen all the first two. This research pieces is new for us, but all the all the indicators and signals are there that there's certainly accuracy there. I mean, threw some numbers on how accurate or how valuable we've seen. You know. sixty pips could be generated on a single trade. As anyone who's looking at finance sort of basis points. But it's new and that's I guess. maybe the last thing I'll mentioned to it like thirty eight percent of a message is voice tone. So if you have new information, You have a higher chance that you know no one's trading on because it's new and that gives you an opportunity to me. You make something right. Make it make it valuable.
Tj_Vantoll:
I don't know if I totally follow that exactly though, because what I'm like with my experience with a models is you, you have to tell it like, if you're building an image recognizer, you have to say
Jack_Herrington:
This is.
Tj_Vantoll:
this is a cat image, so that it knows this is a cat. and then you have to say this is not a cat image, So it knows that it's not a cat. I would think with voice, though like there has to be some level of you have recordings of people like lying,
Paige_Niedringhaus:
Uh,
Tj_Vantoll:
And you say this is a lie or people like confident, but they're not confident. I'm curious. Like what the training process is like for that sort of thing, and how you manage to capture that.
Sean_Austin:
Yeah, well as a great point. there's two lenses or two angles. We went into it. One is emotions. There's tons and tons of label data in the world, like open source scientists or scientific data sets where you can determine emotions. So if you want to be descriptive about this call, Well, actually we wouldn't be the only ones who could stand this up. Are not the only ones who've done it, but actually providing it for the financial setting and on the right corpus of audio we uniquely do. But the emotion part to your labeling or your training quest. There's a tone out there that people interested or have done right, which allows you to think about any use case for audio. And that's valuable for people who are looking at these calls right. So you have an interview with T. J. and you're on. Can you know C, n B, C? You know what's your emotion on it? I can make a judgment as a person, but I'm not, you know, being trained on a hundred thousand labeled sample, So maybe I'm not as good as if the machine told me it was excitement, you know, or confidence right as the as the emotion. So that's more of the. That's the simp, Or one but valuable piece, The predictive nature of what we do, which is a forecast, like a model, A traditional model. That was a very. I mean, that's what our secret sauces is like, a matter of lkeunsupervized, learning, looking at all these financial metrics that exist, so we know we know. historically, For twenty years we have audio files. we know how companies have performed, but performance is like thousands of different variables. We have. You built the models and the measurements all around those. So you think about like you know what is the highest level value? It would be stock price. But if you can dig in deeper you know, Could you predict thirty day returns? Could you predict sixty day, et cetera, et Cetera? So that process, actually, to your point is not trivial. It takes a lot of compute. It's been a lot of basically trial runs, and finding that you know formula where you can say that this actually is valuable audio meas Men and valuable audio tone is why I think a lot of people haven't done it. There was a lot of investment money or eaten amount o invest money, but essentially millions of dollars have been spent to to build those models.
Paige_Niedringhaus:
Yeah, I mean, I'm still trying to wrap my head around how this exactly works. I guess one question that I have for you is you said that there's you know twenty years worth of audio that you can use to kind of learn and start to train the models and things like that. Is there a central location where you can access all this or is it really kind of a cobbling together of? I don't even know where you go to get data this far back to start, Kind of iterating on
Sean_Austin:
Yeah, so that's another great question. So that is exactly what I thought when we started. So where did you get this big audio set? Because if we do it on a year and you do it on fifty companies, it's like you know. Okay, You might as well not do it. Not not interesting. But so what we found is there's so. we have a relationship and have now for three and a half years with a company called Fact set, So facts, that is, you know, think of like a Blomberg competitor. I think that, actually, even as old as Bloomburg, too, but slightly smaller than Blebupbut, huge, now public, multi billion dollar organization for financial services, a big big platform data platform Right, So they actually pipe us this audio systematically, So we acquire it from them. We were the first customers who ever acquire it from them, and they're also the only company ever to offer it
Jack_Herrington:
Oh.
Sean_Austin:
in an enterprise feed, so that
Jack_Herrington:
wow.
Sean_Austin:
that was also a little bit lucky, but we definitely get a nice bill from them, which is worth it. but we get this audio just raw About like M. P. Three is right, is what we pull in, but to your question page, it's a systematic method, so it's not us crawling the web. It's not like you know. on structure. It's very structure, and its from the corpus that people care about, which is mostly earnings. called us, earnings called. So that comes in, and then we do all the layering on top of it, which is T's question right, tons of different training. a lot of call, trial and error through unsupervised learning, and now some higher order things like emotion, classic Ation, all that help people understand these calls in a way where like you know, I listened to a lot of earnings call ease. I almost have to right, but I'm still not
Jack_Herrington:
Uh,
Sean_Austin:
listening
Jack_Herrington:
huh.
Sean_Austin:
to one hundred thousand hours. You know that would be like I forget. Isn't that like years, or think of the years of life or something? So I have a couple of maybe hundred, maybe hours at this point. But what are big pitches is like wudn't. you want? The machine? Who has? We're up to three hundred and sixty five thousand hours is part of our model. You couldn't even listen to that. several people could lift into that thing like lifetimes. You kno. Realistically. So it's a support like a decision support. you know, system here and
Paige_Niedringhaus:
Hm
Sean_Austin:
M. Yeah. Anyway, So that answered the audio corpus question Cause that certainly was a big one for us.
Jack_Herrington:
So what does this look like like if I'm on the other end of this, If I'm the guy that wants to make decisions based on this, Like is it is? there? Is it like a video where it's like I'm watching this co talk and it's like this guy. This dude's lying. This dude's lying, you
Sean_Austin:
Big.
Jack_Herrington:
know
Tj_Vantoll:
Yeah,
Sean_Austin:
big
Jack_Herrington:
or whatever,
Sean_Austin:
Red.
Jack_Herrington:
Big
Sean_Austin:
The
Jack_Herrington:
red.
Sean_Austin:
screen gets red,
Jack_Herrington:
Yeah,
Tj_Vantoll:
yeah,
Jack_Herrington:
rat.
Tj_Vantoll:
yeah,
Jack_Herrington:
who?
Sean_Austin:
starts shaking
Jack_Herrington:
who?
Sean_Austin:
and
Paige_Niedringhaus:
Comes
Tj_Vantoll:
yeah,
Jack_Herrington:
danger
Paige_Niedringhaus:
down.
Jack_Herrington:
zone?
Sean_Austin:
sirens go off
Tj_Vantoll:
live captions
Jack_Herrington:
Yeah, right,
Tj_Vantoll:
like
Jack_Herrington:
Exactly
Tj_Vantoll:
not actor. fact track, false,
Jack_Herrington:
expect half that number. You know.
Paige_Niedringhaus:
Uh, uh,
Sean_Austin:
it. Just it just terminates. The connection actually goes one step further,
Tj_Vantoll:
Uh,
Sean_Austin:
just stops.
Jack_Herrington:
Right
Sean_Austin:
stops
Jack_Herrington:
exactly.
Sean_Austin:
the discussion we
Jack_Herrington:
stop trades.
Tj_Vantoll:
automatically dis invests you.
Sean_Austin:
cut, Cut him off. We, I mean, maybe one day
Tj_Vantoll:
M
Sean_Austin:
it gets incorporated
Tj_Vantoll:
M.
Sean_Austin:
into like, You know, What's the thing I just saw from that T. B. T with Micristoff? Like then, they incorporate in the like the robot now too, like the had that like, Enjoy That was you were
Paige_Niedringhaus:
Oh,
Sean_Austin:
able to talk
Paige_Niedringhaus:
they
Sean_Austin:
to
Paige_Niedringhaus:
incorporated
Sean_Austin:
with the responses
Paige_Niedringhaus:
into being.
Jack_Herrington:
Oh,
Sean_Austin:
and well
Jack_Herrington:
really.
Sean_Austin:
and
Tj_Vantoll:
Sydney,
Sean_Austin:
big,
Tj_Vantoll:
Yeah,
Sean_Austin:
So I think like,
Jack_Herrington:
Oh, my gosh,
Tj_Vantoll:
yeah,
Sean_Austin:
maybe when you're like you know these things are going to be. then you talk to it right. You don't need to type it. That's probably a world where your tone
Tj_Vantoll:
M.
Sean_Austin:
is
Tj_Vantoll:
Hm.
Sean_Austin:
is in there. But so right now to answer you know very specifically, it's a representation of voice, so it's a very non sexy answer. Jack. It's Set of data around. For instance, you on a call would be two hundred and four values of you speaking. So it's a Total data science product, which is, I use the word not non sex, but it's definitely unique. And then what we're doing though, For the emotion classification are trying to really hit like mass scale and address it up as it were. Is the discussed versus confidence versus neutral, Right, Like these emotions or unhappy? whatever you forget. The top ten, But there's a ten emotions kind of agreed upon The world, so we don't make those, but we can now start to say know Jack's confident on this call, so that is getting closer to a U. Because Ten, we can. Basically you show that in cool ways we can show it on our website. For instance, you can buy, you know, like a very light product from us to the web product. You can get a more sophisticated dashboard from us, which we are offering now, or if you want to get blazed, you know, raw data Becaus. You're a data scientist. You get that, which is a more expensive enterprise product.
Tj_Vantoll:
So I want to go back to the data just a little bit. The, I'm curious a little bit on the how, because my thought was, if you gave me a hundred thousand hours of M. P. three files, I would have absolutely no idea what to do with it and I don't know if you can share how much like are you using any tools to do some of this processing? Are you like? like hand rolling all of the processes yourself, and kind of the the idea? My question is, I'm also keeping in mind like a listener listening to this and has their great idea for their great, like a I projects.
Sean_Austin:
Yep,
Tj_Vantoll:
so I'm curious. Any recommendations you have for people that might have like the sort of idea, but are kind of overwhelmed and don't know where to start and where to go?
Sean_Austin:
Yeah, well, I guess I would apply, maybe even generally to maybe product in a, But definitely I can speak to the voice part. To start with, He, the end end of that question to you, like you can validate a lot of this without having to actually consume a whole bunch. I think right, like very standard, almost like product thinking, so I would that that's a real recommendation, which may be is an obvious statement, but certainly like understanding the corpus of audio. even just a couple thinking about who would actually derive intelligence from it and doesn't have to be tonal. Um that I would definitely recommend to kind of go backwards the tool chain. I couldn't give you a good one because it's all blade. Basically like you think about. Like basic, all data science tools Like nothing really audio centric. The closest one is a company called Deep Grammar Investigating Using Now So there, sophisticated a S. R platform. So and there's a bunch of those like you know. Even technically open A I does that now too. but Deep Gram gives you things like they'll summarize a call. So Very much language based, not tonal, anything, so language with tone is important. but for a real recommendation we've evaluated a lot of these types of platforms and they're very competent. They. they're you know entity. They've also raised a bunch of money, so they'll be around, I think, hopefully for a while, but they're just a proficient platform in the space and who knows what the what they could launch in the future to? so that, but for us, we're just starting to look at how to incorporate them to make us better. Most of our pieces
Jack_Herrington:
At
Sean_Austin:
are thing At a like blade, data science
Jack_Herrington:
Amazing.
Sean_Austin:
tooling. You know, even you know, like our scriptingrour Mat lab, or you know just all these different tenser flow items. We've tried a lot. We have some experience with them and we use. You know, we have a big Google cloud infrastructure to process the audio, so that that's why I started with the advice I think is before diving into all of it, which we did do before we built all this. There's a place to validate a lot with the audio. I think it's a blue ocean right. It's very. If you get into tone. it's very new People are doing like Sunday Health is doing early onset Parkinson's detection from voice tone. Because you can't get that from language. You have people thinking about customer service tonality right, because if you're responding to anything right than you sound annoyed. The layer of how angry you are could be like exacerbated through tone, which I'm sure we've all. I've certainly, I speak for myself. I've certainly experienced it and smashed the zero part of my touched
Paige_Niedringhaus:
Oh,
Sean_Austin:
screen to my finger finger hurts. But that those sort of things could like make a lot Difference. We think at scale we don't do any of them right with the finance part. But I would you know, definitely encourage and I try to get into the voice community with a project. Voice Is one of them. There's an open source. What's it? called? It? A Lennox foundation, which I can't remember of the top of my head, but there's an open open voice network. I think Ovonovion is the is the name of it. So there's ways for anyone in voice to certainly dig in. And yeah, I mean, I would. would We encourage Because we want we see it like exploding soon. right? It's starting to happen.
Paige_Niedringhaus:
So are
Sean_Austin:
Uh.
Paige_Niedringhaus:
you only doing English language right now, or you trying to do different languages in the tonalities that come with those? because I'm thinking specifically of like Chinese, Mandarin and Cantonese, and things like that where tone is. So I mean, even the tone of a word means something different, so I'm curious if
Sean_Austin:
Yep,
Paige_Niedringhaus:
you've started to go into that yet, or if you're just sticking with a single language for the time being,
Sean_Austin:
So so our product right now is only the Us. for almost exactly the reasons you said. It's well, it's the language we speak. My co founder speaks a couple languages. Actually, he does speak some. and, but he is not proficient in them except for German and English. So that's a piece of it. The tooling that we're familiar with. Also just like any use. Deep grammar is an example. Everything we've seen is us century, because I think where we are, but also for us, the hedge pones and the customers Are most the Us. based. So all of the liquidity and all the market really is us based. Is I think something like seventy per cent? So yes, there is foreign entities and hedge funds and an asset managers that deploy capital, But basically the calls are in English, but with that being said, we're trying to look at even
Jack_Herrington:
M.
Sean_Austin:
like Open a eye system is when we're looking at on how we can get to those other languages, so those are extremely exciting, I guess partially scientifically, because it's like they'd be pretty cool. but also there's a part of the market. It hasn't asked us. Um, so yeah, we've actually even to the point we've pitched this week or I pitch this week with the Hong Kong Science Technology park, which is looking to bring and help expand or help companies expand into a pack. You now into Hongkong as the kind of initial hub. So it's very much in front of us. You know, we hope that that relationship will progress, but there's a huge. Yeah, there's pretty big opportunity. not as big, but
Jack_Herrington:
I.
Sean_Austin:
it's a huge opportunity in a lot of things to figure out lot of things to figure To like you said To get toned across language because it's not a direct one to one for everyone.
Paige_Niedringhaus:
M. hm,
Jack_Herrington:
Well, I think from the listener perspective they like that's super cool. but when it comes to me as a reactiveloper, who you know is working on the front end, there's been a lot of talk about how with a I, there's big opportunities for us as fond developer is to put like a U. I on this thing and experience on this thing. So now you know you've seen a lot of this. You know you, you struggled with this. I'm guessing you know your customers are like I want ever. You're giving me forty Value. Is what the heck does this mean Kind of stuff? like how do you work with Ui developers? To what are you looking for in terms of putting a face on this thing?
Sean_Austin:
Yeah, I think the most important part is to lie very important question. and now right, like when we personally struggled with Is that's where all the opportunity is right. Like taking, I can
Jack_Herrington:
Yeah, exactly
Sean_Austin:
literally say one person is the data scientist right now for what we do, which is where we needed to start, And you can think about like Well, maybe you don't need to build the engine. Maybe you can use. maybe literally, maybe you can use that, P. T all these front ends built around it. Then you can say like Well is the mode is big or whatever? But I think there's certainly ninety nine percent of Val You to your question on making it consumable. Making an understandable We. I can speak to that for years Now With what we're doing at Helios, right, we know we're We decided to build the engine, So now we spent all this time and think it's the right strategy, but all the opportunities, the front end. So what we've done? I mean, these are small examples, but we've partnered with companies that I can say are actually built and react, and we actually just react on our website. So being such a you know, flexible common, Actually, no kind of the standard right that What we've
Jack_Herrington:
Yeah.
Sean_Austin:
done? so we've partnered and tried to build these dash boards. We've also how our website you know personally for us, is a front end into this where we can say like the confidence is zero to a hundred On this call, Right, and the representation of that which sounds in one way, I guess sounds like trivial, is actually the totally the opposite. The simplification of whatever you use as your data or your you know engine. I think in most products right, so you could certainly build it
Jack_Herrington:
Yeah.
Sean_Austin:
like we did. But that representation is all the opportunity, And I can. I. For us, we have not fully exploited it yet. We're certainly learning and working through how to make this. just like totally consumable. Um, which is all through a front end.
Jack_Herrington:
Yeah, I'm thinking about those those times at Like you re watching a politician talk and they've got like some group sitting in a back room that are like you know, on the right in the left or whatever, who are having a dial right, and they're saying, Oh, I'm happy with this. I'm not happy with that and then literally on the screen there's like this long graph of like they look really like that or they didn't like that kind of thing. I mean, it seems like you could actually do this over time right because you might be confident in showing up and saying Wow, we're a really great company. You know, people are great, but then when you get to like our, It's actually going a ship. It's like you know,
Sean_Austin:
Yeah,
Jack_Herrington:
Like hits the floor. You know
Sean_Austin:
Those use cases. I mean, eventize into your previous question. exposing it to other use cases. Isn't so much now for this back end system. I mean, there's definitely some. It's all about understanding the consumption patterns, like making it understandable making it modern. Making it look great. I mean it's you know. I think it's just a really important point. you know. So we, we personally want to go into those other domains. We know we want to start with finance. we want to get to invest our relations. Incorporates. We think that's going to take re configuration, but the same engine To display. It writes, definitely a little back and work, but we, I mean, just from my experience. it's just so important to think about how someone is going to consume it. You know, Make it just an easy experience right, like easiest, simplest, you know, is always always
Jack_Herrington:
Right.
Sean_Austin:
wins, always wins in the end of the day.
Tj_Vantoll:
Yeah, it's you're what you're offering. Here is a little niche right like, at least I would consider it nice, like just voice tone specifically in like in earnings in a finance world, and it limits the potential little bit, but I'm guessing it also makes it more accurate because like your data set is, you're not consuming things from all different ways people talk right in
Paige_Niedringhaus:
Hm,
Tj_Vantoll:
different scenarios. It's like all the audio is coming from A similar type of input right? It's like a bunch of a whole bunch of earning earnings calls. but I feel like that could help make the model potentially more accurate. It might thinking about that right versus like, if you just took in like general audio of anything that could be in a whole bunch of different form as ways of speaking.
Sean_Austin:
Well
Tj_Vantoll:
and what, not,
Sean_Austin:
exactly mean that is. so we thought the formula for us early on very early on before we built anything. It was. How do you think about high value conversation? Like who's listening? What's being what moves up on it, and then to your exact question, what's the structure? So you'd want systematically. It to be almost the same right which you can think about it. Actually earning Called are very similar. They start with introductions which are very scripted, and then they have, and a sex Ans, which are, They try to be scripted, but are more free form. and that time is actually even almost the same, the most always sixty minutes. So there's this very refined or contained environment around it like a format. I think you use the word, but it's also very valuable as we know. A lot of people tried a lot of dollars after earnings when they hear certain words. Some, I'm sure trading in a non systematic way for the or without having a decision tool. I should even say without having some tonal tool. So they're hearing it with their ears, and maybe they make it All. But if you could now push that thirty ight percent of a message which is voice tone across the board, you'd have like high value structure conversation. What we want to do is get to the other ones like there's a lot of conversations out there. We'd love to love to analyze. You know,
Tj_Vantoll:
In my sort of limited experience with
Sean_Austin:
M.
Tj_Vantoll:
this sort of stuff has been a lot with image recognition, and I know that when you try to build a data set for something that can recognize something in an image,
Sean_Austin:
A.
Tj_Vantoll:
it's amazing the things that can throw it off almost instantly that you don't think about like, even just like the resolution of the picture, Because like you train a model to recognize at a certain pixal density or whatever,
Sean_Austin:
Yep.
Tj_Vantoll:
and then you change the pixal density, And unless you specifically train the mode To account
Jack_Herrington:
It's
Tj_Vantoll:
for that, it's going to be. It has no idea what that is like. gathers nothing from the previous one and the same thing like I flip something over. Well to like a human that looks like. Oh, that's the same thing just upside down. But like one of these models it's like. What the heck is
Paige_Niedringhaus:
Uh,
Tj_Vantoll:
that right? Like
Paige_Niedringhaus:
uh,
Tj_Vantoll:
it's It's like a different pattern being thrown at it. so I imagine there's
Jack_Herrington:
a.
Tj_Vantoll:
that was kind of. the what I was thinking Behind that question Is like some of the consistency of the input can help when training some of these things, because I think sometimes with some of A I processes there's a tendency to have it solved all the problems in the world, like which I'm sure we will probably get to at some point, but I kind of like these more nitch offerings, or at least I think they're good starting point for something to be building for the next few years.
Sean_Austin:
There's definitely
Jack_Herrington:
Eh.
Sean_Austin:
a ton of like you said things that can throw off a call. Just like an example. I got asked all the time is. what if I'm sick? What if
Jack_Herrington:
Yeah,
Sean_Austin:
the person's sick
Paige_Niedringhaus:
I was
Sean_Austin:
on
Jack_Herrington:
exactly
Sean_Austin:
the
Paige_Niedringhaus:
just
Sean_Austin:
call,
Paige_Niedringhaus:
thinking
Tj_Vantoll:
Yeah,
Paige_Niedringhaus:
that
Jack_Herrington:
just having a bad day.
Sean_Austin:
So
Paige_Niedringhaus:
right,
Sean_Austin:
that's where like context comes in. For what will we do quite a bit? Which is like we can't ever get directly into some one's head and know if they're told the memory that was true or false, for they remembered it, or the their future expectations are different. You know. So there's there's all these context. These language is like We always talk about. You want to pair them together because that's how we as people understand it. Even body language is important. but no one really does that yet on earning calls. So there's
Jack_Herrington:
M.
Sean_Austin:
there's certainly
Jack_Herrington:
hm,
Sean_Austin:
a ton of ton of things that can get you snagged up, and I guess people listening or anyone thinking about it right. I mean, I can attest as you're saying to like in the different, similar but different domain, right of modeling, there's a lot of very particular things that could be done. It also almost goes back to the the other question, right, which is are you going to bid Your own engine? And it's certainly very scientifically interesting. It certainly could be more valuable, which we believe right if you own that proprietary piece to it. but there's I think a huge world, if people could just you know, take things that are like what we're sayin with that, T B, T, this whole very powerful engine that now people can re can figure, skin it. think about other scenarios, and all of a sudden you get like this huge. I mean, right, the fastest company to a hundred million users, I think ever took a month or something. It's just a testament it. I meant how U. S can change everything. Um, Yeah, one powerful engine can drive quite a bit.
Paige_Niedringhaus:
Okay, so I'm curious. Let's back up a second from before you were doing this start up. You said that you worked for Spotify and you worked for Lorna and you did sort of big data ish stuff. But how you know this is? This is such an interesting and unique it needs to be in. You know how how do you get started or how do you get the kind of experience that then comes, you know works out for a start up like this, Because it's it sounds So particular, so I'm curious if you wanted to get started in something like this. what would your advice be? or what kind of skills do you think somebody needs to have to do to be able to do this?
Sean_Austin:
Yeah, I would only answer it partially with we, or my. My background almost feel like forced me into it. Because if I'm thinking
Jack_Herrington:
Uh,
Sean_Austin:
of Spotify,
Jack_Herrington:
uh,
Sean_Austin:
I'm listening to music all the time. We were running the largest North American programming so I was. I think it was fifty five thousand minutes of music one year, which is like thirty one days. So it was like the number one on the planet or someone. listen because of my job, But it was that more with their M L acquisition, so they bought a company for a hundred million. The Echo nets, which powers there in Or Echo system now and continues to. And then, Claren was building an a platform to do fraud and credit score, basically across our whole network, So I mean, I guess, for me, I mean that all was very helpful if I was to look back or or talk with someone who is very interested in voice. I mean, I think it really starts with what what you know, and voices everywhere conversations. Even, maybe it's not just voice, but you can think about where understanding can be like, amplified through it. so not even You know Voice tone Specific is a conversational analytic, which there's a lot of now tools that you can get access to voice tones a little a little newer. But that that world is like, I think, just totally wide open. I mean, I've thought about like Alexa, as even to, they're still not totally totally at the same level of like the Ap store, But there's quite a lot of Alexa developers. Quite a lot of integrations you can do with Sere now. Chat T is all conversation right, and that's a whole new world, So I mean, I would start with The skill set would be. I mean, I guess there's more product or customer research or just your experiences on. You know a domain of understanding that you could dig into, And then I would try to start with tooling that is available. I think T. J asked the question to like, I would. I would. if you were new to it, I would not advise trying to make an engine. Just there's some
Jack_Herrington:
Uh,
Paige_Niedringhaus:
Uh,
Sean_Austin:
some pieces
Paige_Niedringhaus:
huh.
Sean_Austin:
there right
Jack_Herrington:
uh.
Sean_Austin:
that if it's new, it would be tough. but there are a lot of resources. Besides people mention some of those communities. There is clearly a trend that Happening, and there's only exploded now with conversation. and if you think either between using like the large language model tools, or like you know, Googles cloud system of Tools or Deep Gram System of tools, there's tons of tons of ways to like again Like I think that was Jack's question. Skinning like the experience starts with. Like, Well, who am I trying to get it to? Is it is at podcasthost, Like if it's podcasthost maybe you can just do a little re configuration and that would be really valuable for him right, so I feel like it's a Yeah. Those would be some good ways to think about it and start.
Jack_Herrington:
Yeah. what do? Yu Tin Would be some easy entry points for folks who are just beginning in their career and they're like Okay. I see, I see the train coming into this a thing, and I wanted to jump on board. What would be some easy off the shelf. A P is, obviously they're not. They dont know how to train me of this stuff. They just want to go and say Hey, look, you know potential recruiter I put some on this. I understand this.
Sean_Austin:
Yeah,
Jack_Herrington:
What do you tink? What would be your advice to that person?
Sean_Austin:
So the three that jump out around around voice gin at toned, but conversation or a voice? it would be deep grams, one of them
Jack_Herrington:
Sure,
Sean_Austin:
at. Not to plug them fifty times, But deep gram is a really
Jack_Herrington:
Yeah,
Sean_Austin:
well established platform. then
Tj_Vantoll:
What
Sean_Austin:
I will go ahead to.
Tj_Vantoll:
when you mention these tools, can you give like an example of the sort of thing somebody would build
Sean_Austin:
Yep.
Tj_Vantoll:
like if I'm building like a simple project E, even if it's just like, It's not like my super amazing business idea, but it's just like something I want to build just to see it like in the rat world, Right, if like, The To do list is the classic example. If I want to learn,
Sean_Austin:
Yep,
Tj_Vantoll:
react, the first thing you'll tell him to do is like build it to do List. If you build a good to do list, you'll know the basics and then you can go from there so I'm curious when you bring up these platforms. If
Sean_Austin:
There
Tj_Vantoll:
you
Sean_Austin:
you.
Tj_Vantoll:
Like an example. Like what is the sort of thing I could do that Like would be my spring board to then build my more complicated thing.
Sean_Austin:
Yeah. So so with a deep gram or use another one? simple symbol. A S. a sort of direct competitor with those guys that that appritthattat like simple, Hello World on this type thing is around summarization and detection of some of the topics, so their a p. I will allow you to easily push audio, and then using their a p, I easily extract. Basically, you know a handful of analytics from it, but building a non
Tj_Vantoll:
Yeah,
Sean_Austin:
black and white cube with the text I think would be a very interes Ting first pass Right because your system is able, you're able to just you know, get all the access to all the pieces you need, so you could easily make. I think a really like for ance Really cool mobile version is what my mind would go to like. You could definitely make a much better version than all the tautorials I've seen or all sample code I've seen of that because you know every everyone's on mobile anyway, So you could think about it that way. the other one with chat, g, P. T would be just interfacing to it, I think, and being able to show you can like construct, Um, like a thread, almost is interesting to me right, like Th. they handle everything for you also, because it's just so everywhere. For now, it's almost. I feel like you have to have some thinking on it. you know, So if you're building an example, you know, working with React on it. I mean that that to me is a very topical one, building conversation threads and again, sort of like you know, figuring out either some existing libraries for how you handle Message threads, but pulling that together could also look. One could look better, but also you'd be interacting with hat t t in a certain way, which I think is, you know, just totally totally needed. The probably the coolest one I think is making a dialogue flow. So Google Clatplatform has dialogue flow, which is a way to make voice apps. So it is definitely
Jack_Herrington:
M.
Sean_Austin:
a little more complicated, but to be like, Do you want to make a Google Assistant app? So you actually need to? So there there is an inner face. When I would think about how to build an inter face to it. You do need to actually quote, Train it. It's not really directly like data training like it's more closer to chat, T. B T. but their system is interesting becase. You could almost show thing in my mind without having done this. We want to. We have an elexaappbut, we use like a platform to build it, But you could. I think very easily, inter, face a dialogue flow, and not only have now a voice like you know, headless where you less version, but the actual interaction to build that To me would be very interesting because it's all a P. S. So you know, I want to almost build like voice at Builder. You could show like, almost like a back office ad man, which I think is closer to your to do list example. But you'd be using basically one of the best conversation. Um, you know app back ends that exist out there and it's super cheap. I mean, I think virtually free because Google's. You know, if you're just starting, all the stuff is essentially free and it's extremely, I think Flexible. Oh,
Tj_Vantoll:
It's exciting stuff and one cool thing I think for our listeners. If you have a Web or react, background is like the reason I think conversations like this are cool
Jack_Herrington:
Yes,
Tj_Vantoll:
is because we as web developers have a little bit of a leg up from some people in that not only can we use some of these a P. s, but we know how to build inner faces around them and often times it's not like when you're using something that's like a generic platform like your cat, g, p. T. Google, Google out anything like that. It's not necessarily the the engine underneath, but it's the unique thing you put on top of it. Like if you can just come up with a really fun U. I of using that thing in an interesting way. Oftentimes you can create something really cool and really valuable.
Sean_Austin:
We are for the websode that I mentioned, we are in that right now, So very practical is making our version of this data the engine right, All this crazy analytic, just in a very consumable mobile, friendly way, we know how to do it right in a way like technically we, we have people in some resources, but I mean, really like distilling it down to a great version of it. You know, it's super valuable back to the one percent versus ninety nine. So totally, I totally agree with you. I mean, it's funny we're doing some of that right now because we need to expand beyond One percent,
Paige_Niedringhaus:
Oh,
Sean_Austin:
But it's yeah, I think with the world that's coming, I mean, there's going to be people who just built models and things like that, but access to it is definitely not going anywhere. You know, people need to still access it
Tj_Vantoll:
There's a real art of taking something complicated. like. Sometimes, almost the more complicated your thing is, the more if you can distill it into something simple and usable like
Paige_Niedringhaus:
Hm.
Tj_Vantoll:
I admire the people that can do that really elegantly Because it is. It is not easy.
Sean_Austin:
Absolutely hundred percent,
Paige_Niedringhaus:
So so can do you think that they're well? Do you do you, or would you in the future potentially offer some sort of an a P. I, that people who wanted to build their own platform or dash board, or things like that could use to consume your data.
Sean_Austin:
So this would be the one direct Helios plug. We do have that, so we do
Jack_Herrington:
Yeah,
Sean_Austin:
have
Paige_Niedringhaus:
There you go.
Sean_Austin:
a tonalupload A P that we're looking. Actually build all of our kind of the discussion. build the inter face to it, so you actually could just go to the website by credit or buy access and push push date, instead of it being a P. But we do have that. we've named it a really cool name Mercury, but it's really around. Like getting a barometer of what people thinking right through their voice tone. We do have one customer on it, which is, really Is a unique relationship with this financial technology company and platform. But it's there. I mean, it's not where you can just go get an A Pike from our website yet, but
Jack_Herrington:
Uh,
Sean_Austin:
we are,
Jack_Herrington:
uh,
Sean_Austin:
and that's why this is. I think a really cool and important conversation is. we're starting to now actually promote it right like we want, and we have now. I want. Say, maybe like three, kind of like Private Private Bates, But we could do a lot. I mean, our only band with is just. I mean, it's Basil. The restriction is just the conversation right now, so it's easy for us to hand off a pike. That's not self, sir, But there's documentation. It's been used for years. Um. it utilizes our engine, so that right, it's like the same engine that can are using in a certain way, and yea, for me, I'd be super exciting to have a lot of a lot of people hammer it even if it's just like fooling around with it and see what's possible. I mean that that is our vision of the future is to be that engine that you can get access to in certain ways. We just needed to start at some place hard in it in one domain before we think felt comfortable that we should offer it up to everybody or Offered. I should say you know, Be confident in mobireoffering up.
Tj_Vantoll:
All right, I have to. I have
Jack_Herrington:
Very
Tj_Vantoll:
to ask a
Jack_Herrington:
medical.
Sean_Austin:
Yeah,
Tj_Vantoll:
very meta question. So you,
Jack_Herrington:
You were.
Tj_Vantoll:
you are a start up, so presumably you have investors,
Sean_Austin:
Yep,
Tj_Vantoll:
and presumably you brief those investors at some regular interval, and
Sean_Austin:
Yep.
Tj_Vantoll:
in a way that would resemble an earnings call.
Jack_Herrington:
Yeah,
Sean_Austin:
Yep.
Tj_Vantoll:
Do you? do you run your own audio
Sean_Austin:
It's always
Tj_Vantoll:
through
Sean_Austin:
hyper
Tj_Vantoll:
your
Sean_Austin:
confident,
Tj_Vantoll:
own system?
Sean_Austin:
which is weird. always the Como.
Jack_Herrington:
Nice, hyper
Sean_Austin:
Now
Jack_Herrington:
confident
Sean_Austin:
it's this new category that only actually shows up on certain calls, but we.
Jack_Herrington:
Right,
Tj_Vantoll:
Uh,
Sean_Austin:
we
Jack_Herrington:
right,
Sean_Austin:
have
Jack_Herrington:
right,
Sean_Austin:
been
Jack_Herrington:
right, right,
Sean_Austin:
asked that. That's why
Tj_Vantoll:
uh,
Sean_Austin:
it's a very. It's very funny, but real question. we've got to ask that we haven't done it and then show the report just because we really offer this blade. You know, blade system right for earning sculls, But I've been asked enough where what I can tell you, which is a little. I don't think I'm in the same As this person. but it's the same example is. we're going to release our review of S. B F. So all the interviews we've done on Sam bankmntfreed, we have. I think it's like a couple hours between pre
Tj_Vantoll:
All that,
Sean_Austin:
pre impost collapse and we've spent about three
Tj_Vantoll:
My God.
Sean_Austin:
three months going through it and all I'll say is not hyper confident. It's a very interesting and I think impactful outcome. but now I've started to think of using. Is there like an early warning signal That could exist from voice tone if you listen to interviews or people talking, or might invest, or calls. For instance, right, if you had the same signal, you know negative, say just negative five on a scale, right, negative five, negative five. Does that mean something over time and we're seeing that it probably does, And that should be a pretty cool thing to. I think, a pretty cool thing I don't know feels like, Feels like the Ight article On a word it correctly, Because it's not trying to predict liars or something, but it definitely would make you. I think, dive in, dive in closer. If you In these, seeing the signals right in these negative signals,
Paige_Niedringhaus:
Oh, that would be fascinating. I
Jack_Herrington:
Absolutely
Paige_Niedringhaus:
would read that
Jack_Herrington:
yeah, yeah, yeah,
Paige_Niedringhaus:
immediately.
Tj_Vantoll:
Yes, sounds like you have somebody that's good at marketing there, too, because that's I could see that being.
Sean_Austin:
I do this when I'm doing the marketing part. So just now I'm
Tj_Vantoll:
Yeah.
Sean_Austin:
but we have. We have a marketing director who is great in a team. but they, we've been asked. I mean, that's where that came from, so it was actually funny. We just asked when it happened over and over and over and we've never done something like that, but it was. I want to say two dozen times. Probably when it first happened, they're like you need to do this. You should do this and investor text me or some one on a customer call. So we said we'll do it and the results are fascinating. So that's how that's like March. We're actually just fine Losing the draft to go to some outlets. Some reporters,
Paige_Niedringhaus:
Stay tuned.
Tj_Vantoll:
Yeah, well, Ishan. this has been fascinating. Is there anything we've missed? Anything that and I also toss it over to Page and Jack. If there's any final questions they wanted to ask, but any topics that you wanted to touch on that that we haven't brought up yet.
Sean_Austin:
I thought this was great and a totally unique lensbecause. I love the, as some one who has a computer science degree and did a lot of web development. It's like one of the reasons we're very interested in doing this and to pages question that a pi is, really, I think our future where we have a lot of builders and developers in front end and back and using it. So just yeah, glad I was able to be on and have a great conversation with her one.
Tj_Vantoll:
Awesome. Yeah, from pages from a web developers perspective, That's all we want. We want an a P. I, K, and
Jack_Herrington:
Yeah,
Tj_Vantoll:
then
Jack_Herrington:
exactly
Tj_Vantoll:
then
Jack_Herrington:
just just
Tj_Vantoll:
we'll start
Sean_Austin:
Yeah,
Paige_Niedringhaus:
Just give me the data
Sean_Austin:
give me, give me that.
Jack_Herrington:
Well handle. let's take it from here.
Sean_Austin:
not even the documentation.
Tj_Vantoll:
well.
Sean_Austin:
I'll figure it out.
Jack_Herrington:
Oh lord, no,
Tj_Vantoll:
that's second.
Sean_Austin:
You
Tj_Vantoll:
Yeah,
Paige_Niedringhaus:
Right
Sean_Austin:
go
Paige_Niedringhaus:
when
Sean_Austin:
say
Paige_Niedringhaus:
I get stuck. I'll look for it.
Sean_Austin:
All right.
Tj_Vantoll:
I'll just ask. I just use chat g v T for that
Jack_Herrington:
Exactly
Tj_Vantoll:
anyway.
Sean_Austin:
Well,
Tj_Vantoll:
So
Sean_Austin:
that's right, Co pilot.
Tj_Vantoll:
uh, all right, why don't we get into Picks where we pick something fun interesting from around tech from around our lives. Whatever, And Jack, do you want to kick us off with picks today?
Jack_Herrington:
Yeah. what was I thinking? I had a really come back around me. Sorry
Tj_Vantoll:
All right, we'll go. We'll go to Page
Paige_Niedringhaus:
Okay, So my pick this week is going to be a show. If you're familiar with the series Yellowstone, It's one of the spin offs from it. It is the, I think, the newest one which is called Nineteen Twenty Three, which follows the same family of people who are in Yellowstone, but in you know, two three generations back where Montana is still quite rural and wild west. S. so it's been really fun. Watch you know. there's more cowboys. There's Indians, there's There's even part of it is taking place in Africa. So if you like seeing some of the African wilderness at the same time, I would say that it's as good as the original and it really is some interesting back story on the characters that you know and love from Yellowstone. So nineteen twenty three on the Paramount Plus channels, going to be my pick for this week.
Tj_Vantoll:
Awesome, My pick for this week. I'm actually going to pick two different apps. Both are appsthat. are offered by libraries. So one is Libby l, I, b B. Why, the other is coupla h, o, p, l, A, and both of them are just apps that your local library probably offers one or both of them where you can essentially borrow books. And what I was interested in is audio books actually just for free. Just you can. If you had to your library, they might have other apps as well. If you are big into audio books, both have both those. Es have pretty surprisingly good players and you just listen to a bunch of free stuff Just as long as you just go in and borrow. It works kind of like any other library system. The library has so many copies. If no one has them, you can borrow them and just listen to them for free. So pretty cool system Jack. You did you find yours?
Jack_Herrington:
I did. I did. Yeah. So last night I was having a conversation with Bang. It was so cool like I got
Tj_Vantoll:
H.
Jack_Herrington:
invited, you know, and and it was
Tj_Vantoll:
M.
Jack_Herrington:
awesome. They've got this cool new chat you where you can just chat with a search engine. And we got into this long conversation about the Star Wars prequels and sequels. It was hilarious and it's great. You know, I really enjoyed. It was my first interaction where I really enjoyed it with an E. It just I did Want to stop talking to it.
Tj_Vantoll:
I just got the email that I was invited last night, so I got to get in. It made me use Microsoft Edge, Though, so I was
Paige_Niedringhaus:
Oh
Tj_Vantoll:
like,
Paige_Niedringhaus:
yeah,
Tj_Vantoll:
maybe tomorrow I'll try it.
Jack_Herrington:
It's okay. It's weird. I ctually. I actually started there, but whatever that's fine. I'm good with it.
Sean_Austin:
Uh,
Tj_Vantoll:
Yeah,
Sean_Austin:
uh,
Tj_Vantoll:
Shawn, what picks do you have for us?
Sean_Austin:
So I'll do. It's like a mash up. but if I can do to because it's related at least the same domain. So one, because I'm
Tj_Vantoll:
Yeah,
Sean_Austin:
obsessed with it right now is the last of us on Bo Max,
Jack_Herrington:
M.
Sean_Austin:
So that is for anyone who doesn't know It's based off a video game reason. It's mostly cool to me, probably cause I actually played that game when it came back away back when, But it's like a post apocalyptic, you know, journey across the U. S. and in no, no spoiler alert, but definitely totally worth watching, And I think it's almost like season finale time, but the other one Which I feel like I have to is a Ap that mashes up voice chatgpthtechnology, In the fact, it's an app right, So it's called a Wasis right now. You have to request those O. A. S. I. S. I think you need to request it through their website and they give you like a test flight link immediately or Android. I think, But maybe it's only, but basically what it does and it's interesting for this conversation because the very light, but the version of it is you speak into it like your musings right towards this Big button To speak. So anything you want kind of like a note taking,
Jack_Herrington:
M.
Sean_Austin:
but it will reform at. It'll output like ten different versions that you can use from like a social media version of it to a summarize executive version of it to a very
Tj_Vantoll:
Uh,
Sean_Austin:
simplified
Tj_Vantoll:
uh,
Sean_Austin:
version of it to a text message like couple of word version. So it gives you all these like ways to discuss and understand it for where you need it to go. And I've seen it. Actually, I mean, I find it helpful already. I've had I for like two weeks. maybe, um, so yeah, I mean, I feel like that Ties everything together So definitely check it out. I hope it's not is only because I feel like that at ll get me in hot water, because
Jack_Herrington:
Yeah,
Sean_Austin:
you can,
Paige_Niedringhaus:
Uh,
Sean_Austin:
giving me to the
Paige_Niedringhaus:
uh,
Sean_Austin:
Android group, But
Tj_Vantoll:
M.
Sean_Austin:
but worth checking out, and I guess if it's not an Android, maybe someone doesn't enjoyed version of it,
Paige_Niedringhaus:
Very cool.
Jack_Herrington:
Yeah, exciting.
Tj_Vantoll:
Awesome. Well, this has been a ton of fun As a last question to you. I'd like to ask like, if people want to follow you, follow what you're doing. What's what's the best place for that sort of thing?
Sean_Austin:
So the best would be probably a websitebcause. We do quite a bit on press, and then everything links out, so it's Helio Life Dot enterprises, So it's the only Dot enterprises you'll ever type in, probably on purpose,
Paige_Niedringhaus:
Oh,
Tj_Vantoll:
I was going
Jack_Herrington:
It's
Tj_Vantoll:
to say.
Jack_Herrington:
a long
Tj_Vantoll:
That's
Jack_Herrington:
name
Tj_Vantoll:
the
Jack_Herrington:
man
Paige_Niedringhaus:
That's legit.
Sean_Austin:
but I couldn't do that dot I o or something for a company. I guess that was on a two cliche, so that's the best spot
Tj_Vantoll:
Yeah.
Sean_Austin:
it has pressed on us. That as our links, I mean, we're linked in heavy. I'm on Twitter and things like that, but love to connect with Any developers front and back and anyone technical, because it's really about like uplifting a community and people who want to join in on it.
Jack_Herrington:
Fantastic.
Tj_Vantoll:
Awesome.
Jack_Herrington:
Yeah,
Tj_Vantoll:
Hey. this has been a lot of fun. so thanks for joining us. Appreciate
Sean_Austin:
Thanks
Tj_Vantoll:
it
Sean_Austin:
for having me.
Tj_Vantoll:
All right. Let's an awesome chat. Thanks everybody for joining us and see you next week.
Jack_Herrington:
See you next time.
Paige_Niedringhaus:
See you then.