Founding Dreamtonics: Creating music with AI Vocal Synthesizers

It's like asking when everyone's using AI for making music do fear that one day will not be able to sell like a SLE guitar or will the drum like the advent of drum machine completely erratic drums which is kind of ridiculous because people still want to play music they still want to enjoy you know the process of making music and I think that's where the value of our software lies in it's now designed to you know just replace a thinger or is

Not completely being used to replace someone we tend to think ourselves this music instrument kraa from De tonics have created a piece of software that can have people well AI sing in your doll you can basically write music with voices and can we think of this as an instrument sure that makes sense what's really for refreshing about this interview is he comes with an engineer

Mindset hey guys and welcome back to the sound connections podcast today we have a different type of episode we have all the way from the opposite side of the world where I am canri welcome hello where are you right now I'm in Tokyo so you know there's a station got Akihabara where know anation and the games where our companies located in the

Center at hard at Cabra amazing and you have an really interesting company uh dream tonics is a company that works with AI but in kind tight of AIT different way that you know most people used to um what we're going to talk about today is using AI as a part of your creative process you guys have created um a few different things but one of the really interesting things

Is an AI singer where you can actually use that as a part of the creative process of making music midi adding uh lyrics and having the AI sing and I can imagine there's a lot of use cases for that and it also places itself quite timely in the middle of traditional production tools and you know all the way to the right ofso udio so I think there's a really interesting

Space to you guys are um doing right now but before we start krew who are you and what do you do I'm really kind of engineer type of founder you know um I started looking to this whole thing called focal since this um 14 years ago while I was still you know in Middle School it started as the hobby project

And just know about large and large and fin accin okay amazing so did you have any ground in music or was it just in enging side I wish I had this that one thing I kept thinking about it you know um you started diving to techn side later I you realize you don't know enough you know music theory and all those things you can't even play instrument and you go back to learn

It yeah okay so that you went on a journey going back to learn it after you sort of like okay there's some interesting things here but I guess I'm too late to become a master pianist or anything you know but I think that really speaks to a really interesting point um the listeners on the podcast who would have listened to a lot of episodes would probably know me quite well by now but I

I do have a strong background of music I have a masters in music but I am more of a director type like executive type and what I struggled a lot with my early days of my career is there's a lot of things I couldn't do because I just wasn't a good enough musician and that really speaks to sort of this next generation of creators in my opinion where um the skill set of the

Traditional musicality uh is probably going to be a bit less than what it was before not even a bit like probably a lot less than what was before and creativity and expression changes its form and shape a lot and this is what um dream tonics also does in my opinion so can you tell me a bit about you said you started 14 years ago but how did you found the company so 14 years ago there was I mean

Probably the history is longer than that there is this Africa Vocaloid you know it's is it's also from Japan it's from Yamaha um it's Sofer to type in the lyrics you know lay down nose and piano interface and it sings in kind of semi- robotic way but it's starting to do totally different um audios I was quite into

That it's just so like has that musing quality to use but I was thinking if I could make my old version of it and make it more natural and how did that go about like did you just say okay this is this is this product and I want to make a better version of it yeah I didn't thought too much you know you know though Oren All Buzz about a things back then it was really simple

Just you know make a sare make a sing okay I think you make it more simple than it actually is but go me through the process in detail like did you have the light bulb moment or did you just started programming your freedom did you have a full-time job and did this on the side what was the context I was I was in I was just out of grade school you know like uh 12 13

Years um doing programming as a hobby making Flash games so like hob game programmer and uh and back then there weren't all this AI stuff so if you're trying to syn this any you have to do the samples you have to sample all the words a human can

Pronounce like you basically say a e and then you concatenate them you know you cut the recording into little chunks and join them together to make up for the wars that weren't there and why did you do that it's fun I mean I didn't I didn't thought anything about how it could be used for

To musician anything but it's just really fun doing that you know you can you can record SS and manipulate them and uh draw them together do some Pitch shifting and turn it into complete different thing so now you guys are a company and you know I was before I talk I was just going through YouTube shorts and there was a lot of people who covered like dream tonics and your products and it was it was really interesting to see the

Reactions like people were pretty mind blown about a tool and I can imagine you would you would tell me better but I can imagine there's a lot of uh creatives out there that uh a songwriters to some degree or producers that has great ideas but have a hard time putting it out into life what's your reaction you've had to your approach so far um I think my Approach it's

Really um empirical I'll say try to figure out what works what doesn't work and uh it's not like we have this amazing a technology we're going to apply it here to make it sound Supernatural but more like we want to make something that South N could be useful and later I realize it could actually be useful and let's see what our best choice you picking technology to make that

Happen and uh it got more serious you know um around 2017 when I was in college uh I went to University of Illinois and but still you know snowballing my hobes project and uh I started experimenting with different ways to model the voice ising Fus your network and at some point it worked

Really well so I decided to start a company and just deciding to start a company is not something you just do like what's your background there like it do you have entrepreneurs in your family why was that a natural step for you no I don't I mean uh the push factor there okay is like I just spent so much time into making this work at one point I realized

I couldn't graduate and yes sorry go on yeah and there is the pole Factory you know it's like if I get this work it might actually have some commercial volue um back being was nowhere near as close to the Quality we get nowadays but it's on the verge because I do have some well I don't know music I do have some musician friends and I see they're struggling with working with really

Updated technology um basically all what you're seeing like 5 years ago right people like all the plugins the sampls um like the technology used to model all sound not just vocals but guitar like piano they stay back to 20 30 years ago Dees old technology I see there's a chance we can actually make something decent something

More updated work all so where does your motivation lie does it lie in the Curiosity about the product's potential does it lie in the impact like what's your driving force on continuing pursuing this I'm just super intrigued by the engineering complexity really it's just the engineering complexity of this word

Um it's like it's like rocket engineering because in order to make a good starting vocal synthesizer you have to know little bit in all things and to know like the basic is signal processing like how you manipulate the signals or what the consequences are mathematically and then you need to know a bit of like Michelle Ley on the stat the stat

The statistics about how you know um you control the voices and how your voice compareed to the human voice and then you need to know a bit more on the psychology side uh we call Psycho Acoustics how people perceive sound and also vocal production you have to know bit about the you know all organs and what happen inside your body and how it control the sound how it you know fter

The sound it has based everything it's so interesting and we'll go much more into the sort of the details of that but I do want to understand so in 2017 he said okay this might have some commercial value and I'm probably not going to pass college because I use too much time on this right and he said you wanted to start a company um but you know there's many ways you can build a company you can bootstrap it you can build it yourself you can get teammates you can get

Financing whatever what was your journey in sort of scaling this to a commercial viable product by that point I actually already have a sort of working product so I learned this many years later actually um but I didn't F into you know looking for investment or anything um I look at like what other people are doing you know what they called there's amazing

Thing called a startup like you just start a company like a startup right so I called myself a startup and then I realized years later that I'm not actually a startup because there's no one funni it I was just selling things yeah so did that this grow organically or did you have investors come on board at one point uh the beginning of it um of it was pretty much

Or all organic and then what you know you put so much work into it um there's already something working so you're not worrying you know how in all Source I mean there's nothing lose it's just one company right you just have to find way to sell it but that's actually a harder thing yeah that is incredibly hard so how did you do

It so you look at the market back then um there is not a single like practically uh usable uh vocal synthesizer in the market um you ask around a producer friend you know no one's seriously using it for producing the I mean there is some way like if you make D music you can sample chop it and

Perhaps using some creative way but nothing like replacing a lead sing or even use rambling so meanwhile in Japan like talk earlier this a vocaloid there's a whole subculture based on Vocaloid where the naturalness of the voice is really focused but rather it's there something you can associate with it's like creating a virtual singer identity that

You can associate s with so you mark it really hard on that particular identity you could have like a character design like a Japanese animation character um and uh the voice Bank itself or the sample Library itself is really just an addon it's like a collection item with voles Associated to that character so people will just buy it as

A collection they probably wouldn't even use it but I mean that's one practical way you can sell it back then right if you don't have the super good quality so you basically said there's an existing niche market of fans in this area and the way that I'm going to sell it is I'm going to create characters that they would like to collect pretty much I mean in the

Western world is pretty n but in Japan it's huge yeah and again that's you know you know I don't I don't know the EA market so well but that's one of the things I hear all the time like having this um it's called it a Community engagement around things you can immerse yourself in like universes personas that kind of stuff is a huge thing yeah and it's interesting so you

Basically tapped into that way of thinking in order to start selling a product yeah so that's why you know I flew to Japan trting to company here okay so you flew to Jaan Japan to start did you have any relationship to Japan before or you just said like this is the best place to do it um Japan is a common destination for international students um I mean in the

Mid of Pacific right um so uh I did full there a couple times during vacations um during school breaks so there was a musici hackathon taking place in 2016 or 2017 I think um I joined that you know without too much purpose just to one of my friend tag me into it um and then from there I learn

There's a whole Community doing things interesting things like this so I always had find maybe one deck I sell products here but how did you finance the journey until then do you have like a regular job or no I was a college student what it expecting no okay cool amazing okay wow okay so how did you take it from this

You know niche market uh selling to these two to have where you are now for example like there's a few steps between then and now like what was the journey in between there uh you have to start somewhere I mean as a business you have to sell uh I mean if you go down the you know the Boost wrapping Road um you have to sell first so no matter what Market

You're in start n is okay um the first thing is you really have to start a character business but I'm not the kind of person to you know really Del into the animation and stuff uh I watch a lot but if asked me to create my own I couldn't do it so you had to find Partners you have to find the company actually does the character itself and has to interest him pairing it with a voice and uh they also have to because

Physical distribution in Japan is so huge you have to be able to print all the boxes stop them with CDs and burn all data into it you know and that's how you went about it yeah so we were looking for you know Partners in Asia first Japan China and unfortunately not there were a few company started with that from the very beginning um they made their all characters I think we made like five 10

Of these to actually get it, you know, picking up speed and being able to support ourselves organically. And how specifically did that work? So we're just trying to understand, there's these companies that would like to sell voices of characters. What we uses using this for, uh, from the customer perspective? Yeah, first collect it and the second um

You know, it's like having an AI companion but it's not really AI is able to see anything you want to seeing. So you could have your own Silly idea, you know, maybe it's an original, maybe you just want to see like your favorite sound being covered by your favorite character. But it was people who had like Daws like, you know, there were music producers that would be using these products. Actually about half of them

Another half of them are just complet beginners, uh, they don't even think of, you know, in terms going pro and making music, but more like it's just an additional way to interact with their favorite character. Amazing. Okay, it's totally I guess Unthinkable to a lot of like our users nowadays. And so who are your users

Nowadays I mean the animation thing Japan is stillo huge, but uh, I think we have a lot more music producers who actually use our software for practically, you know, making a track, making a lead or harmonies or, you know, creating demos um showcasing it to their customers. Because I think that's the interesting thing and we can sort of dive a bit into it. Like I do a lot of uh I call it research, technically it's not

Research, I just play around with a lot of different tools all the time and, you know, I'm always tacking along on the new stuff every single week, every single release I check them out. And my conclusions are that there's basically this, there's three directions that right now is. There's one that is, you know, traditional music production like most people would know. Then there's

This uh fully autonomous where the intervention of human is extremely limited, but of course you can steer promps as much if you're good at it, so you can get the output that you want. But there this intersection where um people are still inherently creative where they want to have customizable ways of saying what they want to say, but they want to use tools that can make it faster, more convenient, less expensive

And that's sort of where I would place you guys, which is a very um timely thing to do. I said this at the start of the episode, so I can imagine the users that you have uh would be creators that for some reason doesn't necessarily have access to a singer or that they sit in scenarios where they would love to not pay for a singer come in and do things, they can create a lot the creative stuff

themselves and also have a better um income Associated what they do is that correct that's like these semi-professionals professionals uh to some point but uh I like to make distinction that uh in fact how is including how it's being used it's not designed to you know just replace a singer or it's not completely being used to replace someone um what

Tend you think our this music instrument. Yeah, and I agree, that's a really good sentiment because it's an instrument that you can call it a demo, you can call it an outline, you can call it support Dev elements. Let's say there's is a great singer but you want to do, I've seen sort of YouTube with your products like doing like a gospel choir in the background, like things that will be, you know, either one really hard to do like in a fake scenario where like you're having a

Singers do love different vowel sounds or just um not possible financially. But here you have this supplementing element where you can actually create this environment that sounds real, that can support the Creative Vision uh by using tools like yours. So what's your take on this whole thing with, you know, EI generated music versus traditional music production, where do you see the

Futures moving? Uh again from our point, we're really a music instrument maker. So it's like, it's like asking, you know, when everyone's using AI for making music to fear that one day will not be able to sell like a SLE guitar or, you know, or will the drum like the Advent of drum machine completely eradicated drums W

Which is kind of ridiculous because people still want to play music, they still want to enjoy, you know, the process of making music. And I think that's where um the volum of our software lies in. I think that's so refreshing and also a bit provocative saying that, you know, creating a musical instrument, because you know in reality it is, it's a vocal chord uh where you can use the tongue and mouth to articulate

Certain sounds that means words um what's been the been the reaction from the industry, like the traditional music industry in this, like have they been supportive, concerned, like what's been the reaction? I think the general senent is actually very positive um we even have users that didn't think of us as an AI, so that was really interesting discussion. User by ask what you think

You know, sinv really is and they say it's not, it's not AI, it's not doing anything, you know, I'm supposed to do um uh I guess part of it liing controls the user have in their hand um if you use like a Tex to sound generator, it has a sort of a style attached to it, sound like kind of generic or it sound

Really like someone, but sometimes you can pinpoint who that artist is, sometimes it just feels really representative a certain gender um but in our case the resulting sound like the style it has really depend on who's using it. Yeah, I it's just very interesting to have that perspective and I really haven't thought about it that way before

But in reality that's what it is, it's an instrument that people can use um but what do you think is going to be the future of consumers in this space? Are they just going to go towards these kind of tools where things go faster and more convenient? It really depend on what you want to do. I mean thatly true is that uh um a lot of probably the h huge

Percentage of the music we're hearing these days, probably it's a brilliant percentage are, you know, being mined for really just the economical value of it, like the background music. You could go to Starbucks, you could hear some background music um these kind of things, you know, these are really just data drip and you can just generate a lot of them. And I think there's no doubt people are

Just going to use it sound generator for that. Yeah, but it doesn't mean, you know, everyone's got to listen to it, you don't want to activ to listen to this kind of music. What I think is really interesting, one of the things that I think a lot about is sort of the value of Music in general um if you come from music industry uh and you would understand like the concept of Market size in music industry would also very quickly recognize that

It's sort of undervalued uh when it comes to like the impact it actually has and a societal even economical level. There's a lot of research that goes into so what's the actual impact financially from music in the surrounding Industries and workplaces and it's quite enormous, but the size of the music industry itself is quite small um I'm really intrigued about sort

Of the character thing, like your first sales approach, like building into that Niche, because one of the really interesting things with music right now is how we can, it's not a great word, but like democratize the way that people approach music, building music uh we have this Professor that's was on a podcast called Danel Roland who's also um head of strategy for Lander and he said uh

Which I really like that um he loves to play basketball but he doesn't have an ambition of becoming a professional basketball player and the way that AI is enabling people to use and inact and create music in a different way doesn't necessarily take anything away from the professionals who do this full-time. So what you're saying indirectly, which I love, is when you

Started saying you found this indirect monetization channel that has something completely to do with music but was a consumer group that wouldn't interact with music in the traditional way that market does now. Do you see more of that happening in the future? It is already happening, this exactly what many of you are pointing out, like you said there's a, I think building upon your idea, there's always the, you know, the volum of the

Music to the listener versus the volum of music to the Creator and uh it might be true that on listener side it's a bit depreciating, but on the Creator side there's always people who just, I think most people are still just enjoying the process creting music and they want to put their own personality into it. Yeah, whether they're processing while a human singer or not. Yeah, so what's the future of your

Product? Are you going to like open up for people doing their own models of their own voices or like what's the future there? Um I think there are certain demands, so we actually have paired with a different product right, so we're like in terms of product design, we're expanding the family like so there is a s v for generating vocals and then there is another one called vocal flx or transforming

Vocals, so there is a possibility, you know, building a whole pipeline of these tools to really accelerate and uh and I think not just accelerate but also improving, like making it possible to make new kind of vocals, maybe not just vocals but, you know, all kind of sounds we can make. I'm we get bit back but there's one thought I have while you're talking which I think was really

Interesting is you said in 2017 or, you know, when you started this, that it was basically just sampling, cutting up words and now you're uh AI enabled or AI company. But what I think is really interesting is that the concept really hasn't changed, just a technology that gets you to I can assume quality faster. Like what was the transition for you guys like, when did you see okay we could use certain technology to make this

Better? What was the experience of that transition of 2017 to now, 7 years later? Uh that important transation actually was pointed out by our users um at first it's just experiment, you know, uh we decided to maybe, you know, we've been doing it with characters um until that point but maybe we released one without a character and see how he was going to react to it. So

It was still the standard way of making the voices, but we just decided not to use a character with it, but just rather um you know where we have the display, you know, like a little thumbnail for character with putting two alphabets and uh it took some point to um some time to pick up, but uh we're actually inde seeing, you know, professional musicians using that

Product to make their own demos um that was huge confirmation saying that the technology is getting mature for outside of, you know, our animation business. And where did AI come into the picture there? Is that in create creating that non-animated figure? Yes, that was also around the time um the AI technology become mature

Enough to be used um basically if you look at the progress of technology for AI, there is one thing that specifically the speech scientists are worried about, that is um before like 2016 or so there was a problem called like over smoothing problem um what you do is when you train an AI model to generate the voice like

Given uh music score lyrics, people doing solving this test until that point was trying to frame it as an optimization problem, so they're trying to teach the model to generate the best answer to a certain M score. The problem is there's not just one answer, there's like a million ways to seeing the same score. So if you're trying to teach it explicitly using that method, it ends up learning

Like the average of the all the answers it could have, so it's going to sound BL which is called over smoothing problem. Yeah um but after that period there was like transition, someone found a method actually to be able to model like all different kind of variations given the same input and that's I think the point when the technology really got mature enough to be used, you know, to recreate that kind of that kind of, you know,

Random laser Dynamics within like what humans can do. When you say founded technology, is that you know, I'm just trying to understand because I don't, I don't fully understand, is it the same way you would say that some mathematician found like a equation and that something changes how people calculate stuff? Is that the same thing that happened? Similar, but you know, mathematici are always like 50 or 100 years, you know, ahead of us. By the time

We F engineering fining solution, they're probably already dead. Yeah, okay, yeah. But okay so that's what happened, someone found a way to um train a model or work with models so it basically had a different way of looking at how to get quality output. And what happened, like you read a white paper or you saw technology, what was the sort of the moment where you said okay this is something I want to pursue, like how did

You discover this method? It was a gradu discovery, it was, I think it started was in 20 maybe 2015 or so Google released what would called uh the wave net, like the first T speech model is able to since this Crystal Clear speed and then people just follow and take it AP and see what is part actually make it work and then make their own

Variations, but uh I think the very beginning run super slow, takes like three hours to generate one second, but then there a lot of work while we make it faster. Yeah, but what was your Revelation? Was that like in 2019, 2020, have you run for three gears and then you say okay this is the same method we can use? It took a lot of iterations and not

Just not just iterating technology sell, but you have to build a whole infrastructure to make it work. Now another very important part of thinking is um you know a lot of these like AI sound generator what so whatever they run on cloud is not something you can just download and run your own machine. Yeah, this hard because first you don't have the infrastructure to run the

Large models fast enough uh by infrastructure I mean like all the codes you have to, you know, R coding a very specific way um to make things fast and uh many cases they just not enough like the building blocks for an engineer to implement them um people always thinking building something on the server nowadays, the other is how do you

Know control the like how to protect the intellectual property of your own models, so how do you do that? So to get that point uh we have to basically R all the wheels um that process alone took me like three or four years, that's before even releasing our product. It's amazing and I think really interesting about

This space is there's no really quick way to just get a solution for this, like proprietary technology or things you ground up takes so much time. And I try to say to Founders, this is my opinion, I might be wrong, but I think I'm right, is there's two ways of building AI companies right now. I believe there's the proprietary way uh where you build something yourself and there's also two

Directions, there's one you should have started a long time ago and you just spild a lot of time building it, or you do something that has a very strong BTC focus and you get a ton of money to do it, but like that's the two ways. Well, the other one is utilizing existing technology through, you know, wrapping things or apis or whatever to create very impactful knee use

cases. So there's like, that's really three, like one to the one side, two to the one side that's proprietary, and one that's basically in N use cases. Um, so you basically belong to the school of what I believe is the proprietary technology that just take a long time to develop, and but of course scale also matters. So did you get investors at one point? Like what was the, what was that point? I think there's one thing special

about doing audio. Um, probably true some point to other field, but I think audio is a like a prime example of this, that um, it's nich enough but it's not too nich, right? It's n enough that you w't imagine like Google or like OB nowaday they going to now like some they just randomly hit we building s with pair with powerful interface, you know

um, but at the other point it's not like, it's not like you can really skill it up like by adding just, you know, a whole team of a Engineers. It's a very technically, it's a very nich field to get something working decently. Um, 2080 rules still apply here because 20% effort copy someone else model gets you 80% good, but really what the pr musations are care, they even

maybe not Pros even, they really care about the last 20% or 5% H, and to get that level quality you need a lot of time to build up knowledge, to build up understanding of the speech production Pro process H. So did, do you guys have a team? Like how does, how does your company look now? We do have a team, although for the

first few years really just onean business. Um, I had a fortune, had a time to really dig into all these tiny details, learning myself the hard way and make a lot of mistakes. So do it again, you know, from today I will say it's going to be really hard. Yeah, oh yeah, 100%. What's really interesting is

there's two aspects to this. Um, I believe a Founder should have a passion for what they're building beyond the financial return. Um, and I see a lot of founds doing this mistake where they get caught up and all the potential, and I think that has its Merit to some degree, but I think the reality of a Founder's Journey, I've been a Founder for 12 years

and there's many things I've failed with and you know where I just fell flat my face, but I'm still doing it of course because I love being an entrepreneur and it's nice to build something of value, but really it's because it's the passion for building, like it's the creativity, and that's what I see with you. But what I think is particularly interesting is it's your love for engineering it seems, and one of

the things that I try to comment very heavily about the music industry is the needs to be opportunity for people that does not love music in its own insulated maybe not so fortunate way to create value in an industry as well. And often times the industry is a bit get G kipped from Outsiders because like, oh we have what, our way of doing it, you know

this is our rules, this is our Frameworks and this is our network. But it's not a really a good thing because it's not a strong industry per se. So having people like you where the passion lays in engineering and still providing Val to the industry is quite important. But how have you experienced the music industry? Has it been okay to work with, or you just been doing your own things so you haven't noticed? Like what's been your journey

there? They're gradually building up, right? So in the beginning I know nothing about MYRY, I was just interesting solving my own problems. So you see this is the thing, um, most engineers they just went noes down to solving their own problem and they don't mind how people's going to use it. Um, that's not the right mindsight, but it took me some time to actually figure it out myself. So at one point you know, uh, you realize when you

actually want to build a product for people to use it, they have to understand what they're using it for. So in other word, try to teach myself some music. So um, I actually got an interet A like really just a recording shop, my local record in shop, and stay with them for one month and that's what they do. Um, you know, try to go do some of their sessions, um, you know, do all recording

geers they use Pro Tools, that's where also I Le Pro Tools, um, and you know work with all the Mac Ms and all the beers and very important with the people top to the people here their struggle. It took some time to build up all right, because I can imagine that the creative field and the engineering

field, even though it's both about creating, a lot of times I think the personalities can be quite different. Um, so for you transitioning from being engineer just really interested in solving your own problem to being aware of how to solve others must have been a personal challenge as well. Like that's getting outside of what you know and what you come from. Has that been

conflicting for you? I will put all of under engineering framework. I'm not just engine in a technology, I'm engineering a company. Yeah, okay, still problems, right? But uh, now it's a turn into all out Warfare where you have to consider, you know, not just engineering how technology will solve your own problem, but also how the company will work together to solve your customer's problem. And then you apply

your engineering way, you make assumptions, you test your assumptions, you test it with the different ways and figure out which is the way that works the best, and then go back and all the dur SES. Amazing. So you didn't necess try to, you know, fully deep dive into their culture and their way of thinking, you just TR try to understand it enough to have an engineering approach to solving

their problems. Oh I mean, um, understanding their culture is definitely important, but it's important under the framework of engineering the whole business up. You know, if you want to sell one way like you know, trend of thought linking to the solution is you have to understand where you're at, you have to understand it so well so you can base all your assumptions out, or you

risk making assumptions out of nothing, which is very likely to be the pth failler. Yeah, from a business potential standpoint of your business dream tonics, where do you see this going? Like do you see that you guys would be acquired for your technology someday? Are you building towards to stff getting a big customer base? What's the long-term Journey that you

foresee? You know you ask the thing and I don't really know what I'm saying is I just find it fun to work with. I mean, I really love what I'm doing. Um, I love making the software, you know, making a software that people's going to use and make it helpful, make it useful to all people. I enjoy the process, so not really considering selling it

anything, but I'm definitely going to grow it, definitely going to grow it. And how does your team operate right now? Like you as a CEO, I would assume you still do a lot of engineering. Like do you have guys helping you run the company? Like do you have more Engineers? What your setup right now? Yeah, um, so goes back to many years ago when I was, do you know, um, doing as Hobbies project

I did have a few connections online, just real close friend making a program together, and uh, later want decide to start a company and reach back to the am, that's how I do with my engineering team H. So is that how your team is right now? It's a lot of Engineers or there also operations, or try to sort of paint that for me. We have Engineers with

the sales and uh, really important thing is work in different countries, so starting in Japan now, um, going back to China and then the Western Market, so there are separate teams working different things. And how do you navigate being a CEO? Like you know, it's different than just, you know, building something alone for years. It's um, it's one of things you

realize you have to do because you want to overall make things to work, you want to overall sell more products, and then in my engineering mindset, what is the problem is blocking it? And we don't have enough people. What you do? If you going to Har out people, if someone has to do the management, so it's you know, it's a very logical result if you think of this way

you need someone to, it's a business. Um, I guess the question is whether it be me or someone else, but as of right now it's you. Um, and you know, being personal on my own side, so I also run a company and in our portfolio we have around 20ish people and then directly we six employees in like the main company that I

run. I think for me that's the most difficult thing to do. Like I'm a, I would find myself as a very weird mixture between an autist and an engineer. I'm an audio engineer by trade, but I'm also extremely artistic, and one of the things that I find a lot of issue with is I get so hyperfocused on the job, like the product, the

exploration, that for me to navigate the human complexity on top of that is just very overwhelming H. How do you navigate that? Is that because you hired the right people with a culture that aligns with yours, or like what was your approach? I have to admit that this hard, it's hard because for me it's just too easy, you know, to spend too much time on engineering side. You always want to push

it to the extreme U, Focus too much on product and also you know your product too well, but um, I think a lot of it goes to really just talk to different people, hear the thought, figure out the struggle, you know, um, and really the motivation behind the struggle. Hey, and how do you, how do you sell the vision internally? Is that we engineering

something amazing, like is it we doing something great for creativity, like what's your, what's your internal talk when you talk about your product? I think the first thing is everyone in business, they're like users themselves to some point, so they're really interested in the product itself. Okay, that's really the most important thing that keeps everyone

aligned, that they're just really interesting. This maybe they don't come from necessarily a music um perspective, they could be more engineer, they could be more sales, but Aina gr is a very interesting product and they love it. So what do you see this uh from a more macro environment? So assuming out from your company and from the

music industry in general, like a company like yours that you know in traditional ways are not a music company, you're um a company that builds a product that's can be used in the music industry or the people uh in it, but what does this mean for the music Ste in general? Do you believe that technology companies are

going to have a lot of impact on products here? Uh, do we see that the music industry is becoming less insulated? I know it's a big question, I don't know if you can answer it, but like how do you see this developing from a more macro environment? I don't think it's found change too much really. Um, technology has always been a pushing factor for music media. I don't think I need to explain it, but really you know, going from analog to

digital and to, and if you talk about distribution, you know, all way to streaming. Um, for us um, software company is not too different maybe from pluging company, but uh, if you're trying to be the best software company here, then I guess the most we can do or what we should do is to change people's workflow, is to that new workflow is actually useful. I

mean for people to make uh songs, I mean specifically sounds with vocals. So someday um, your first choice into making the song is you want to first um, at least create a demo, you know, using a voice synthesizer, and from there on you can, you know, uh, decide whether you want to keep that synthesizer or you want to hire real singer. But with think at some point people they're just going to

default into, you know, maybe I'll just spend five minutes to see how this replica self liking a synthesizer and then I'll decide from there off. Yeah, I think that's really interesting and I think the fear in the music industry has for some time been like the job displacement and everything's going to change, but what I also see more and more is what

you say, sure technolog is going to disrupt things, it always has, and especially it's going to move away around jobs, but also seeing more and more that a lot of these companies that supposedly is the threat of the music industry seem supplementing to some degree now, and companies like yourself that comes very much from the outside looking in, it's just like, well this is a tool we wouldn't be able to make ourselves, like we didn't have as

an industry the persons the enough um technical Focus to provide this ourselves, and we might have had an idea of what we needed, but we weren't the people to solve it, and that's probably what history has done over and over again, but I think it's easy to forget it, and also like my generation under, like I'm 31, like we haven't been a part of this huge digitalization that changed

everything, like you know, for us, for me, from when I started music in 200 uh, when 10 or something like that, whereever you went into it really hasn't changed much from 2010 to 2023, like it's pretty much been the same. The only difference is plugins will become a bit better and computers become more strong, that's about it. So now the first time with youngish people where we see this

fundamental change, so we get really worried and confused, but I think you really hit the um the nail on the head, like this is, this has happened before, you know, it's just history repeating itself, but yeah, it's different, but still it's not really different. And also if you think about you Mission something like it com from the AL busit right, um, it's not always like you know

um, you're from a more advanced culture anything, you know, but it's not like in the in the precolonial world where all war is about a stronger, you know, culture it dominates over other culture, but in this case more like with begin, we begin as an outsider into music business, but then AR to improve a product, we learn a hell

loud into what producers do, what they like, what they don't like, what their problem, and in that process we got assimilated into part of it. So I think one of our um, one of our core ethic is you have to understand the culture first and be part of culture, and I think that's what you're do with the podcast now, and thank you for opening up and telling about

Your journey, I'm really interested to sort of follow you. Like I follow this space a lot and I haven't, there's very few companies in your space that actually does well, um, but I think it also speaks to the complexity of it. You know, I think also people get used to things like releasing all the time, like because of these big Mega pushes of BTC propositions, but in reality what's happening in the background are these

companies that's been working for 7 10 years where suddenly maybe uh Computing has become cheaper, uh new technologies have become more accessible, uh New Concepts have been released in white papers and research, whatever, that makes these companies been bootstrapping or really working for many years can now work faster, um, and I think there's a misunderstanding of where the Innovation lies. It's not this reent

thing that just happened. Like most of the companies, and I know a lot of companies in space that does well, has been there for a long time, long before the hype was there, and I think that's the interesting Dynamic to see. And you know, if I would make a, let's call it a more negative comment to a lot of STS out there, is that there's this illusion that this goes fast. Like I just have this idea, I just find some random AI engineer and I'm going to build this

like no, that's probably not what's going to happen. And I think where stories pick out as well, like anything good where you do something completely new, it's going to take time and effort and you never really know. Like you could also have been unlucky, like this whole AI movement would not have happened and you would still be struggling with, you know, a technology where it would be much more difficult, um, so there's some timing

element to it as well. But of course your opportunity only came because you had done the hard work before that. I think I'm really lucky if you look, you know, retrospectively, I did the right thing at the right time and I was really lucky to be in the right market and lucky enough to realize I'm not just building something just for myself. That is, that is luck. And I think, I think of course, you know, luck is a

weird word, but that's for any entrepreneur. Like the resilience of time over time is just incredibly important. Like I'm right now empy Ventures, we're building multiple products where I'm fully convinced with my industry knowledge, my network work, with who I talk with, that some things we building is going to be great, but the timing is not here and it, I think it's going to come soon, I hope so, but I know, I know

I'm ahead of my time and my vision, so our job is just keep on building, keep on finances as much as we can, keeping bootstrapping, and then when the opportunity comes I know we' be ready. But that's also a weird way of building it because it's, that's the thing about technology, it's just like you got to take a bet. Like this, you know, product Market fit, the traditional thing, like sure it has Val validity, but not

necessarily right away. I guess I just said while things you can do is really just try to amplify that little bit of Chance by doing long enough until it works. But I guess not part of the story is you have to be really good at sales and pivoted to the Right audience to be able to sell it. Yeah, and that's key and that's where I appla you, because as an engineer

obsessed by the product being able to do that and recognize that's, um, that's not easy. This has been a great talk. I really appreciate your thoughts. I think it was a very fresh air into what sometimes in this book has becomes a very much singular perspective from the music industry, so I really appreciate this. Um, what could you give of advice for Founders in this space right now? And I'm

asking you because you've, I wouldn't claim you have seen the future because very few people do that, but you are willing to take a chance on something that was very unclear. If you were to talk to Founders about this space right now, what's going to be your advice? What should they focus on? How should they think? I think the hardest thing, um, is realize your own bias is to, you know, understand what

you're doing from complete Outsider point of view, but if you, if you become too involved in, you really just get, you become blind and you lost other directions and that is very denter thing to fers. Good, I think that's very true and I often become myself happens gr when you're working alone or you have a team that everyone saying the same thing, you know,

everyone is saying H this the bad thing we can do. Yeah, Camu, it's been great to have you in the podcast, uh, and thank you for taking your time. We there are two different time zones but we could make it work. Um, cam good luck with your company and I'm looking very much forward to follow it for many years to come. Thanks so much