Ladies and gentlemen, I'm full of optimism roseline theory of relativity. We're still seeing him quite well. We're growing equals mc squared oh man I created about the future innovation and growing growth in the coming up. This is finding your frequency with your hosts, Jeff spinarak and Ryan treasure each time to speak up, share your voice and hear from the thought leaders.
Ladies and gentlemen, welcome back to finding your frequency. I'm your host Ryan treasure. And it's an exciting day because it's finding your frequency Fridays we got a great guest for you guys today. You know, we always like to talk about interesting topics and you know different things last week we were talking about self care and some different things that are important help kind of further people's purpose and passion in life. You know, we talked about you know, really paying attention to your own body and making sure that you know, you're you're able to operate at the best possible frequency for yourself. So that way you can be the best earner for you and your family. And you know, always being able to take care of yourself as important before you can take care of others. So that was last week's show. If you guys didn't get a chance to take a listen to that, please do we had a great guest on with Robert Allen, from NDC, solutions calm. So go check that out. If you guys want to check us out on social media, you can follow us at finding a frequency dotnet on the website. And of course, you can find us on social media at finding your frequency net on Facebook. And then of course, you can follow me at a radio Ryan one just about everywhere on social media. today's show we're gonna talk about tech, technology, entrepreneurship, and some really cool new technologies that have been, you know, really catching some steam over the last several months. And our guest is a absolute expert in these areas. And so we're going to talk to Kirk Marple who's the CEO and founder of unstruck data, which is a new company that's building the industry's leading unstructured data warehouse for automating data preparation via metadata enrichment, integrate compute and graph based search. Kirk has over 25 years of experience developing media management pipelines, leading DevOps and venture backed companies and structuring successful exit. He holds multiple patents and industry awards and has truly established himself as an industry thought leader. So Kurt, welcome to the show. Hey, thanks for taking time, I know that, you know, you guys got a startup going on. And a lot of the work on those companies, you know, come with, you know, the startup action and getting funding and, you know, series A's and all that fun stuff. And, you know, we can talk about all of that, and what you guys are doing over at unstruck data. But before we get into, you know all of that what you guys do unstruck data and talk about some of my favorite stuff, like machine learning and some of those kind of things. Let's just kind of find out a little bit more about you and who you are, and how you found your frequency in life and in business and how you ended up where you are.
Yeah, thanks. I mean, it's been a it's been a ride. I mean, I started off. I mean, going to college for computer science, me, but originally, I honestly thought I wanted to go to culinary school. So kind of original path. Didn't really follow. But it's, it's been a really interesting, I mean, kind of building companies building software, really kind of finding that that path through life is it's been pretty interesting.
That's awesome. So did you end up getting your degree in computer science?
Yeah, so did my undergrad in computer science and worked for several companies in the Washington DC area, and then headed to grad school on the West Coast actually went to grad school at University of British Columbia, and then ended up at Microsoft for I guess, six good years, had a great time in Microsoft Research there and worked on a lot of I mean, really, future looking products knew back in the day, it was mean 3d virtual worlds. And I worked on Windows Media Player for a while. So it's a great I mean, great experience to cut my teeth after my master's and then ended up starting a company after that, that I mean, it was an almost a dozen years of my life after building it, bootstrapping it, selling it and working for the buyers for a while. So that was a big chunk of what I've done in the last last several years.
That's awesome. Microsoft. You know, I was just using a Windows Media Player earlier today. Oh, really? Yeah. Yeah, I don't like the new player on Windows 10 I use Windows Media Player still.
It's crazy. I mean, yeah, it's I mean, that was a I mean, years and years ago that I worked on it back when CDs were still a thing. But yeah, it was really interesting, interesting project.
Yeah, that's funny that you brought up CDs. So I Limelight as a DJ, playing turntables and mixing records. I used to do a professionally when I was younger on on am FM radio many many years ago. And but I still I still dabble. And what was funny is I had that found all of these CDs that I had lost that were in the back of my closet in my off my home office. And I was just like, I gotta rip all these, I need the mp3 files so I can use them in my, you know, digital, digital DJ software. And I spent three days ripping CDs with Windows Media Player, because I tried some other software and it just didn't work as good as Windows Media Player.
Well, that's funny. Yeah, I mean, I worked on the base, what they called the CD music database. So is, I mean, figuring out like, who, who the artists were and I mean, it's all common now with Spotify, and all that, but that's where I really got the first experience with, like, what we call metadata. I mean, the data around I mean, who, who the album was, and I mean, who the artist was, and all that and worked on, I mean, honestly, that's, I mean, 20 years ago now, but it's honestly been a path that I mean, I've not intentionally followed, but it's, it's always been kind of there, of dealing with all the different file formats, and metadata and media management and all that is, it's been a common thread.
Yeah, and you're still dealing with metadata with your new with your new stuff with unstruck, cuz I know, like metadata, you know, for me, especially in broadcast is extremely important, right? I mean, when I'm, when I'm loading up audio files and our automation, you know, it's very important that the metadata is there, so we know where it came from, and all that kind of stuff. And then I'm starting to notice too, there's lots of metadata that's included in podcasts and the delivery of those podcasts and, and then also, you know, metadata in a lot of new industries that I'm seeing that are really starting to break out, which we can definitely talk about, which is I'm sure, right in your in your wheelhouse. You know, as a lot of companies moved to, you know, the industry, industry for auto kind of methodology of automated, you know, factories, and some of those kinds of things, you know, you're really reliant on metadata to make sure that all the machines and all the things that are being automated, and the AI and machine learning are all, you know, able to kind of talk to each other in those spaces. So tell me a little bit about how unstruck data is using metadata enrichment and all that stuff for your integrated computing your graph based search, which I think is really cool.
Yeah, no love to I mean, it's, it's pretty interesting. I mean, so I, as I said, I mean, I, I spent a dozen years or so in the media entertainment industry. I mean, we had adult software that we'd sold to a lot of the major broadcasters and ESPN and Fox and people like that. And, I mean, it's metadata in their, their realm is, is really used for different suits, and different case, it's, sometimes it gets passed on to like YouTube and Netflix and kind of flows with the content, as well as helping you search for the content. But what I actually found is I worked at General Motors for a while, and then a few different other companies after after I sold my first startup. And the it's a lot of the similar data, it's images, it's audio, it's video, but they're actually trying to figure out what's what is the data about, it's not for eyeballs, like for consumer like, like you and I, it's really getting plugged into machine learning or getting plugged into different line of business applications. And I started to realize, all the tools that I built and been familiar with for the entertainment industry didn't exist in in a like sort of what you call heavy industry, like manufacturing and chemical companies and things. And so I started to speculate while while I was at GM, and after that maybe a lot of the things I built before those concepts could apply to industry. So that was really the kind of seminal part of starting unstruck. And it, it was kind of an idea that kind of started germinating after I left GM and I ended up getting recruited to a couple different companies is like CTO or VP, and still had it in the back of my mind that this idea of building a knowledge graph from the metadata. And using that to organize the data and then search it, and then also visualize it and integrate it with other things could be really interesting. So I ended up at a company in the bay area that I was CTO of, and we were sort of in the space dabbling in this space. They were a drone company. And I started to realize, I mean, look, I think there's a more generic kind of platform here that could be possible that could apply to a lot, I mean, wider swath of folks and so ended up I mean, COVID had something to do with it. It was kind of like we were kind of changing direction on that product. I mean, really wanted to keep moving forward and this ended up working on a seed round of funding, and was lucky enough to get a great set of investors and kick this off. Back in I mean, I guess it's only been about four or five months, really, since we got fully going and have a team of almost 10 people. While heading in this direction. I mean building buildings, whole architecture, and we're getting ready to ship in about, I mean, about six weeks or so.
So what's an unstructured data warehouse? You know, I, you know, I did a show? Oh, I don't know, it was a few weeks ago that we talked about data centers, you know, basically just just how the fact that data centers power our daily digital lives, right? Like, if you're on social media, you know, you're on a website, whatever, like all of those are going through a data center of some sort. But those are all like structured data centers, right. And so what's the difference between an unstructured data warehouse and a structured one?
Yeah, I mean, in a structured sense, it's a pretty classic. I mean, think of almost like a spreadsheet, you got you got tables, or rows and columns. I mean, there's a structure to it that I mean, everybody can kind of have that. He would say that the lingua franca that of like, everybody can communicate over that data in a pretty standard way. And the term sequel is a is an API in in sort of a language that you can query that data in very standard way, in a structured world, but for unstructured data, and, for me, unstructured data is kind of any, any sort of document or media file type that lives as a separate file. So it's, it's what you and I would have as, like Google Photos, or iPhoto, I mean, dealing with your images and video, simply, that's unstructured data. And what we do is we pull out the structured part of it, because there always is some level of structure we can find inside the file. So we have what are called parsers. For all the different data types, we pull out, things that we find useful, like when the time the image was taken, or the GPS location that the video was captured, or, I mean, there's a lot of data in there even, we're dealing with some special thermal infrared cameras that you can have the temperature of what it's looking at, in inside that file. So you essentially have to crack the files open, extract data from it, and then create another set of data that is somewhat structured that you can query on. And so that's what we're doing is kind of creating data from data, creating structured data from unstructured data. And all you have to do is basically put the data into the warehouse, and we do everything else for you. We extract it, we allow you to search on it and visualize it. And that's really the big difference is, I mean, we're kind of chewing up that data that the company or companies give us and we give it back to them in a form that's more useful.
As a founder or business owner, you know what it's like to run your entire business from your inbox between the sales, recruiting and fundraising emails, things can get messy, really fast streak is a CRM designed to help stay on top of each part of your process and your inbox without leaving Gmail amazing. Let's not leave Gmail let's just work from right there st gives you tools for email tracking mail merges and snippets saves you time and you can scale up your email efficiency. In just a few minutes. You can also set up pipelines right inside your inbox. start tracking your contacts and emails through each process. streak helps you collaborate by sharing emails and pipelines with team members. Whether you work in an office out in the field or on a remote team. pipelines are completely customizable. So you can track processes and details specific to your business. access your pipeline on desktop or mobile app to add and share informations and meetings at job sites or however you work on the go sign up for streak today@streak.com slash frequency and get 20% off your first year of their pro plan. Their most popular option that streak comm slash frequency for 20% off their pro plan streak.com slash frequency. So can I referring back to what I was talking about with industry for Dotto and I know you guys working with machine learning, we actually have a show on the network but by SAP who's all about industry for Dotto and, you know, automated manufacturing and all that kind of stuff. How does how does what you do fit into that space of autom automated manufacturing and, and all that kind of stuff? Like what what types of metadata is being used, you know, within machine learning for that type of environment?
Yeah, it's interesting. I mean, a lot of a lot of the companies we talked to have solved the, quote, capture side of the equation, so they know how to capture images or video and they have the devices around or they have people that are really going around with mobile phones, capturing data in their environment. So they've, they know they need to and it's really more of like an inspection use case they know they have to visually inspect like conveyor belts or I mean we're dealing with some ports that are looking to see like a wind at the bumper on the waterside fall off because the boat hit it too hard. And if they can capture the data just via images over time, we can then track that and Through the computer vision, we could say look in, I mean, the first, like 50 photos you gave us, we saw this bumper there, the computer vision, like machine learning algorithm, found this bumper. And for some reason, like, in January, that bumper fell off. And so the computer vision, no longer saw a bumper, and we can look at that trend and sort of see those edges and see, oh, wow, this used to exist, it doesn't exist now. And we can tell the customer like, Hey, we identified a sort of trigger point in your data, that we're seeing a change. And that's a lot what they mean they do all that man. So I mean, literally have people walking around their facilities, I mean, hours a day, I taking pictures to capture the data, but also essentially manually inspecting, and manually having to I having to triage and identify the problems, well, we give them as a facility to automatically do that. Now, it's not a one size fits all solution, you still have to customize the machine learning algorithms for each use case. But we're Our goal is to kind of give them 80% as a platform to build on. And then they may have to tune it and refine it the last 20% for their specific organization.
And I'm sure probably most of the organizations all have kind of different parameters of how you know, they need their data delivered to them. Right. So you guys are structuring that data based upon whatever the customer is looking for. Wow,
we have some common like time and what we call geospatial sort of like on a map view. And then the metadata are things that are common to any customer. And so we we are kind of opinionated that we map everything to one common set of sort of data that we can query on, but then there's always going to be kind of custom bespoke stuff for each customer as well.
So as you guys are ingesting all of this, you know, unstructured metadata to put it out into a structured thing, and you're parsing that data, which I'm which I'm very familiar with the, you know, idea of parsing data, just working in Brian broadcast, right? Because, you know, we take raw WC three log files and parse all the data out for our stats for our radio listeners, you know, looking for, you know, how, you know, what, what was their get method and all that fun stuff? And, you know, what did they access and their IP address? What should we then use to correlate to geolocation and all that fun stuff. But you know, that that type of of parsing and data manipulation and movement takes a lot of computing power? What are you guys doing to manage that portion of what you're doing? I would imagine that it takes quite a bit of processing power.
Yeah, I mean, we're, we're, we're fully a cloud based solution, we actually run on Microsoft Azure today. The one thing we've done is it's, it's called what's a serverless architecture. So it's all based on events and triggers. So we have, basically each file that we see from a customer triggers a set of workflow steps. But the good thing about it is on the cloud, I mean, you can actually burst out capacity very easily. And also then kind of wind down that capacity as needed. So I can run, say, dump, like 100,000 images into our system. And it'll sit and I mean, it'll peak out for, I don't know, half an hour or so, an hour chipping away on all those files, however long it takes, but then it will kind of like wind its way down and say, Look, I leave one or two servers, or one or two workers sitting here, I don't need to have the 1000 that I spun up for that hour. And it's a way to manage costs. But it's also a way that we can do things faster, too, because we can scale out in parallel. And so that's intentionally we made this into a very scalable solution, because we have some customers that are talking to me about, I mean, hundreds of terabytes of data. And we need to be able to support I mean, even just getting the data into our system takes a while just copying data in so we can't take a huge amount of time other than that, because they're already having to give us their data in the first place. And then that does take me take a little bit of time.
Yeah, interesting. Yeah, that's exactly how our systems work as well. We're all in the cloud. And, you know, if, if we're crunching stats, and then it decides that it needs some extra computing power, right, it'll automatically spin up some other processing power. Yeah. And then we also have that also on our on our application side as well, for the front end of our website where, right if you start getting a huge amount of traffic that the server can handle, you know, it's going through the load balancer and the load balancer goes okay, we're gonna split this traffic and, you know, spin up another, you know, virtual server over here and yeah, that's definitely I think the way that a lot of companies are moving towards with technology based workflows, and you know, also ensuring them on the the, you know, the limitation of having downtime to write because you Always have, you know, some type of a virtual machine to failover to in the event that something happens and you know that mitigates I think a lot of downtime too. Are you guys seeing that as well?
Yeah, exactly. And I think it gives us that ability where we can do the failover, we can scale out on demand. And it's a much more resilient architecture. And we wrote it. I mean, I wrote a really, over the last couple years, I mean, in tried different things and kind of settled on this architecture, which I mean, has served us really, really well. So so like I had most of the, I guess you'd call it the platform kind of back end functionality. I've been working actually on a podcast, discovery platform to kind of enhance and enrich the data that's within podcasts and build a knowledge graph from that. And that's what I essentially pivoted into, into this architecture. And we're just using it for for more of a commercial use case, rather than a consumer use case.
Well, we might have to talk about that later offline here, because that sounds like something we could use.
Yeah, I mean, that was that was my original goal. That's I mean, I mean, we still have the IP it's, it's more we arcading taken more of a b2b model instead of b2c. But I think it's, I'd love to come back to that someday, because I still think it's valuable. It's a, it's a stay at a different more consumer friendly use case.
Yeah, no, I agree with you. And yeah, I hope you do come back to that after you get, you know, all this off the ground, because I think it's something that's needed in our in our industry, for sure.
Yeah, no, and I think it's, it's, I mean, it's, even from a research standpoint, I mean, trying to I have a daughter who's in graduate school and trying to research things and correlate different types of media, from video and images. I think it's this whole Knowledge Graph concept. I've been really interested in the last, I guess, three or four years, and there's just so many applications of, of how to correlate data. And we're just really scratching the surface at this point.
Yeah, that's right. So are you guys using, you know, artificial intelligence in anything that you do? Or is it? Is it just machine learning? Like, what, give us a little insight into that? Because I know, those are huge button buzzwords, right now with AI machine learning, you know, a lot of large scale companies are leveraging that I know SAP has their, you know, their, their, their Leonardo AI. And, you know, you have, you know, IBM Watson and all that stuff. So what are you guys doing to leverage that technology within your infrastructure?
Yes, we're, we're leveraging today, Azure cognitive services, which is a bundle of different computer vision and natural language processing, and machine learning API's. And it's really cool. I mean, you can give it a document. And it'll extract the words out of the document, and then look for entities in the documents, you can find like people places things in a document, we can give it images or video even. It'll identify objects in the image. So I mean, you could give it, I don't know, like a construction site, it'll know that there's a, like a truck there, and a building there, and dirt and all these kind of things, it gives you back like, a couple dozen, we call tags from anything that you, you identify. And then even we do audio, like text extraction, you can give it a zoom meeting. And you can actually extract all the words and find out the entities from that. So the idea is we're extracting, essentially, this structured taxonomy, a set of tags, from documents, images, videos, whatever, and then creating edges between them. Basically, I mean, tying together those tags and the media, say, look, here's a company name, I heard somebody speaking in a meeting, I saw it in a document, I literally saw this on the side of a building in an image. And that's where we can create those edges and automatically build up that knowledge base. That's really multimedia. Truly,
Wow, that's pretty cool. I'm on nerding out over here.
I mean, it's, I mean, I think the the other part of this is, I mean, we're using commodity, ml and AI today, I mean, we're using it from Azure, it's just out of the box. But we do anticipate, and this is our plan for later this year is to build it more of like a plug in model that companies might want to plug in their own algorithms, or there's other vendors that we can find, because we're not gonna build everyone. I mean, there's so many different ones that are tuned to different types of data and different vertical markets and things like that. And, and we just see it as a platform, like we're coming out of the box with something that gives you kind of that 80% like works pretty well in a generic way. But we know there's going to be a lot of tuning and a lot of special things that happen in the like, I know, like a port might have different, like computer vision algorithms then, like chicken manufacturing plant or something that we've talked.
Yeah, that makes sense. You know, sometimes just struck me with what you were talking about with, you know, you said that you could feed the AI a zoom meeting, and then it would convert Put it into text and then you could get the tags out of it. This is something that I run into as a challenge all the time when when looking at like transcripts for of audio content, right that people use for blog articles and other promotional mediums and social media and stuff like that. How accurate is the the the model for converting that into text? Because that's what I always find when I'm looking at, you know, these companies that claim that they can convert, you know, audio and video to text and then you know, you try it out. And then it's like, wait a minute, it's very clear audio, they're clearly speaking regular English. Nobody has a southern accent or an East Coast accent and, and then you look at the document, it's like, I got to spend 20 minutes fixing this because it didn't it didn't extrapolate it right? Are you guys? Do you guys find that problem as well with some AI when converting, you know, content to text?
Yeah, yeah, I mean, it's a common and it's not a perfect world yet. I mean, and who knows if it ever will be. But the good thing is I've talked to, and there's new vendors coming out every day, there's a couple, a company called primary.ai company called assembly.ai, that I've been talking to that are kind of like, better than the average. And so like Azure, Microsoft has a bunch of models that they have, but you can always find a specialist. And so another one is, it's even by vertical. I mean, like healthcare has very different language requirements, and even written and spoken then, like a normal industry and stuff. So that's something we do anticipate getting into is like, we hopefully can release more kind of tuned. models, as we understand like, okay, like, here's the sort of language that a manufacturing like a chemical company would use versus a port, or a, like a university, we talked to them about their, their, like building maintenance and stuff like that. So there's gonna have to be tuning that's done at that level to be really, really accurate. But we're really showing the kind of a bit more generic, like, here's, here's the capabilities to get off the ground. And then we know we're gonna have to refine that in specific verticals later.
Yeah, totally agree. Funny. Funny story, though, right? So I brought up a Word document, right? And you know, how you can you can dictate to a Word document. Yeah. So just for, for fun, I decided to try this the other day, I took my microphone, and I put it next to my speakers. And I started playing a radio show off of the website, and I, I'll be done it. Do you know how good the dictation was for word? Yeah, I was, I was pretty surprised.
And then I mean, they've gotten really, really good. And even some of these other companies I've talked to you, or like, even an order of magnitude better. And it's I mean, it's surprisingly good. And so in other I mean, even for us, like, I mean, we want to identify specific things like, like an equipment ID number, or something like that. And I'm sure that's going to be very specialized to what our customers dealing with, but that's an area we see is having the ability to say, look here, give us your list of all the potential equipment, and then we can tune the text algorithm to look specifically for those IDs. So you can make it very, like specialized for the customer with a bit of extra work. And so that's, that's something we're definitely looking at for the future is to I mean, make it more and more accurate, because that's where a lot of the values.
Yeah, I can see a whole bunch of applications to for that, including, like inventory, right? I mean, if you if you wanted to take inventory, and it was just as easy as like, let me go in and take some, you know, snapshots of my fleet, or whatever, right, and then send them through the algorithm or whatever that you guys have. And then you give me a, you know, an equipment list back just based off of me taking photos, that saves me a whole bunch of time. Plus, I'm probably, I'm probably better off doing what's in my wheelhouse, rather than walking to each truck, and writing down the number and the make and the model of the vehicle. And you know, all those things, too. That way I can hand my, you know, Superintendent an equipment list, right? Being able to take photos of the equipment as you walk down the road takes about five minutes, and then just send it over to you guys. Right. I mean, that's, you know, in a sense, I know, it's very low level. But I mean, in a sense, that's one of the capabilities of the technology that you're building and when and and one thing that that could solve, correct?
Exactly, yeah. And I mean, it's an area we've talked, we talked to one company, there's a Electric Company, and they were taking pictures of the IDS on the power poles. And they're trying to identify those and then correlate those numbers on like those metal tags and the power pole back to essentially an equipment ID list. And that's really where we're heading is to be able to automate a lot of that and you could say, well here give us your connect us up to your database as your full list of what we should find. And then we can show them like, Hey, we found like, I mean, give them almost like a histogram of here's every piece of equipment. Here's how many images we found of each piece of equipment and then look for the negative space of like, oh, weird. Like, why did we never see these two, like polls over here or something, and, and also tracking it over time is really important and on a map, so you can then map all of that and say, oh, weird, why is there is this outlier, we found, like, four miles away that shouldn't be there. So that's really where I'm really interested in seeing is starting to look for anomalies and look for trends. Really, for this quarter, we're focused just on getting the data in, and then we'll really start looking at, like, just tracking those insights that we can pull out of it.
Yeah, that's pretty cool, I can see this market being pretty large for you guys, especially with, you know, getting into, you know, civil municipalities. And, you know, some of those kinds of things. I know, my neighbor works for the city of Phoenix, and she works in the traffic department. And that's one of the things that she has to go do is go out to, you know, traffic lights, and make sure that they're working appropriately, get their ID numbers and all that kind of stuff. And she's doing all that manually, you know, with a laptop and a truck and inputting all the data in manually so that can bike maybe give her out?
Exactly, yeah, that's really dead on the kind of stuff we've looked at is, I mean, helping augment, like, it's not gonna replace the human, but it's going to augment the human and just take a lot of that mundane work. I mean, we've talked to somebody at a port, that is literally just walking around six hours a day, taking pictures of things, doing their rounds for maintenance, and things. But when we're actually going to build a mobile application to help augment that, and literally, by the time they get back to their desk, we've already auto indexed and organized all that data and hopefully, triage things. Maybe it kind of bubbled into the top of things they should look at.
Oh, that's pretty cool. So they just take the photos through the mobile app, and then right into the database, they go and that's pretty cool. Exactly. That's awesome. Yeah,
so we that's a that's what we're working on right now. And you can kind of take these sessions of data, and then kind of like, you could say, like, have a manager, sort of watch the the people and then it'll sort of auto triage that data and say, Look, here like I mean, manager person, just take a look at these, these images, because we think there's there could be a problem there. And that's sort of auto triaging that via the, like, machine learning part of it.
Yeah, wow. So many implications. I know that you guys said that. You just got your funding, what's kind of the roadmap for you guys as you work work towards, you know, getting additional funding? And, you know, start, you know, in? I guess, what am I looking for? streamlining what you guys are doing? What's kind of a roadmap for that?
Yes, we're, I mean, pretty much from start to finish, we're looking at about a six month process from getting our funding to getting into the hands of customers in a sort of release state. So we're trying to get to what we think of like as a one Dotto in about, say, 668 weeks Max, we're really just mostly finishing out the front end user interface. At this point, I'm just fixing bugs, tweaking design, filling in missing pieces, we have at least I don't know, almost a dozen warm potential design partners that we want to get using the product that we've already been talking to. And that's really the goal. So this quarter, I guess, up until the fall, it's really all about just getting customer feedback. Not even in a paid way at first, but just getting getting people using it getting experience making any changes that we need to that we may be missed. And then the winter is really more of a bigger rollout. I mean, actually getting people paying kind of getting into more of a normal customer. Yeah, what do you say cycle, sales cycle and then funding wise I mean, we actually have a I mean a good amount of runway still we got 20 plus months of runway already. At this point, it's we've been really frugal, it's a I mean, a pretty tight team, we all have worked together before. And but most likely look for funding next next year, probably next spring, is what we're looking at right now. But we just want to show progress and show customers using it and get that traction. That's that's really what it's all about. And I I've been talking to, it's still talking to potential investors, and it's been really positive, they understand the I mean, what we're doing and what the need in the market is. And we've had a lot of interest in so we'll do a series A at some point, probably early next year. And I mean, really, and then that's the point where we then we start really pushing on building on machine learning algorithms. And I mean, stretching this this out into a much, much wider space.
Wow, that's cool. I really like that. So, um, you know, as I'm thinking about the implications of your product that you're working on, you know, is is is this product, something that could be leveraged for, you know, DevOps teams as well to help organize all of the data points that come with a given, you know, website build or some kind of technology builds
That's an interesting point. I mean, it's technically at its core, it could be, we've started off with more of an opinionated model of really tracking this more geospatially. And temporarily, like over time. And so we're focused a bit more on almost like a, it's a, like an outdoor use case, I mean, things that happen in the environment and are tracked over time. But technically, like I said, I started with this podcast discovery idea that the engine that we built, the Knowledge Graph could actually be applied to a lot of different areas. And we actually just got an inbound lead from a guest, he called them a medical company, and it's somebody doing bio biomedical and stuff. And they need to build knowledge, graphs between like patients and drugs and different things and actually be able to build out a knowledge graph from all the data they have. And it's not our sweet spot today. But like, technically, if we wanted to, like spin off a new product line, or something like that, eventually, I mean, we really could, because the core of what we build has a ton of different applications, not just where we're headed right now.
Yeah, that's awesome. Because I know and I'm sure you understand this too, as as, as a developer to like, when you go in when you go in to start, you know, project managing for a given project, and you start sitting down and, you know, okay, what are my sprints look like and delivery and, and, you know, starting to figure out what codebase you're using, like, all those kinds of things get to be almost almost as much work as you know, when you start coding and actually doing the work in the first place, that the planning, the planning is just as just as hard as the as the actual, you know, build as far as I've seen, and some stuff that I've worked on. And, you know, I think that that'd be cool to be able to have a tool where you can plug in all your data points and organize it for you versus, you know, take out the human error element, you know?
Yeah, and I think even just, I mean, we we deal with this every day of just crossing the lines between design and product and engineering, and oh, wait, did we say this in a meeting, like, I mean, going back and correlating the notes from a meeting, and even what was spoken on a zoom meeting, back to slack and back to I mean, JIRA, like, tracking the tasks and all that. And I think it's, it's kind of interesting, I mean, if we can come up with a kind of a, an insight generation engine that works across a lot of different use cases, I mean, we could pull that back out and apply it to like DevOps and developers or I mean, academia, or different things like that. And that's kind of where I mean, I'm, I've always been sort of a platform guy, like, I always think, broad, but then you got to sell specific, more vertical. But I think the platform that we built horizontally has been a ton of future applications. And it also gives us an ability that, if we're off the mark a little bit, we can pivot a little bit, because I know the technology can support it.
Yeah, I think that's probably an important business decision, especially in today's kind of environment where, you know, we just, we just witnessed, you know, for the last 18 months, or whatever, you know, I don't know how many people started pivoting, you know, in their businesses to keep up with the pandemic stuff that's been happening. And, you know, I know, that's something that we had to do to we had to switch not change our business model, but just, you know, pivoted in another direction with some different options for pricing and different options for, you know, the, the, you know, changing the barrier to entry, to working with our company and doing some of the things that we do just because of the way that the industry and landscape changed because of COVID. So, you know, knowing that that's a possibility in the future for you guys, and having a plan of attack. And knowing that that product can pivot in many directions, I think is a very good business choice. Because, you know, I would especially when you when you talk about business, right? You always have contingency plans for you know, fire and hazard and all this other kind of stuff, but I don't know about you, but I didn't know our company didn't have a contingency plan for a pandemic. No, no, I mean, so now it's like, it's like okay, well now we know that that's possible. So now we have to have a pandemic you know, preparedness
you know, the other trickle down things like my old company, I mean, we I mean, when we were looking at the original product, we were looking at oil and gas as one market segment, but then when COVID did I mean just the whole energy market and everything just got up ended? I mean, are you think about like, supply chains and everything like that, it's when if you're selling into one of those areas, it literally were economy and travel and hotels and everything, like you have to be able to pivot quickly, because your whole market just disappeared overnight. So I think I mean, I have to think about it from a technical sense, but also a business sense of, well just kind of hedging your bets. Like if the if for some reason we don't get attractions one area we do you have kind of a plan B, Plan C plan D that I mean we can we can try and execute on without blowing up the company. I mean, we're just kind of turning the wheel a tiny bit and just aiming at a different spot.
Yeah, no, exactly. I mean, you're not doing a 180 you're just doing a course correction. Yeah, I was in the Navy. So you know, we're gonna we're gonna cut two degrees. Yeah, yeah.
You know, it's I mean, it's that's the kind of way I look at it is I mean, you got to it's you got to be really pragmatic. I mean, you can't put all your eggs in one basket, but then you can't be completely generic, because it's really hard to sell a generic platform to so you kind of have to, we look at it as you got to have to balance both those those opportunities.
Yeah, no, I totally agree. And you've done this before. So you kind of have you know, that experience with the other startup business that you've done. And, you know, you're moving into this one, do you find it easier as an entrepreneur to build this business than you did when you did your first one?
Yeah, I mean, I have to say, I mean, I think it's, I, I went through a lot of ups and downs and pain, because number one, were bootstrapped. So it was I mean, down to our last dollar, and then we'd get like, 50, grand check the next day. And it was just like, ups and downs, and ups and downs. And so in my I mean, my family and the kids were smaller than and it's just like trying to manage, like, I mean, a family and startup and all that kind of stuff. It's it's a huge stress and just huge challenge. And I think now it's just, I mean, life's a little the kids are older, it's less stress of dealing with that same time. And, and just I think, I've learned so many lessons, I mean, painful lessons, but learn a lot of lessons of I mean, how to manage people properly, how to kind of look ahead while managing the day to day and all that. I mean, it's it, but there's also different problems. I mean, now I have investors to answer to right before I didn't, and there is there's a lot more risk, I mean, I have 10 people's livelihood, and their family's livelihood sitting. I mean, on my back, and just making sure that this company is successful. So it's, as we're gonna, I mean, I am anticipating this company to be an order of magnitude larger than my last company. And that's going to be a much bigger challenge to I mean, a lot more at stake.
Yeah, I like what you said that you made a lot of mistakes. I call that failing up.
Yeah. It's so true. I mean, and even now, I mean, I look at I mean, daily of, like, Am I getting too in the weeds with one thing, and I don't know, I gotta pull back in one area and just let my delegate, like, delegation is so key. And yeah, just trying to manage that, like while maintaining control, but not meaning maintaining too much control. But I have a great set of people. I mean, we've all worked together at different companies. And even though we are I mean, completely remote First, we do know each other pretty well. And so I think that that helps. But it's also I mean, it's a maintaining a remote culture is also I mean, important. Just make sure communications good, make sure there's open lines of discussion, and you're always gonna run into pain points and things like that. But I mean, just trying to be open about everything.
That's my love hate relationship with Microsoft Teams.
We I mean, we're a slack company. And you we've used teams as well, but it's, it is it's one of those things. I mean, we got caught in a situation earlier this week, where we're literally trying to talk about one button in the user interface ended up being like a 20 minute discussion in in slack. And if we just like said, Nope, nope, stop, jump over to audio or get on video and show it through, like, you just got to use the tools for the right situation. And it's too easy to get right hold and just not have optimal communication.
Yeah, you know, I have a rule with both text messaging and actually any kind of messaging, right? If I have to send more than two sentences, right, in a in a text message, or like on, you know, teams or something like that, I have to send more than two by more than two sentences to say I'm picking up the phone and calling you.
It's smart. Yeah, no, I mean, it's, it's really true, I think. And it's so easy in the moment to kind of forget that. And it's like, I'm kind of reminding myself of that piece that happened this week. But it's it, you really have to find the right path. And and make sure and also, I mean, a really interesting point is everybody's a different kind of learner and communicator. Like, I'm much more of a visual communicator and visual learners, some people are audio. I mean, some people are better writing, some people aren't and having to kind of learn everybody's strengths and weaknesses, and then communicate in the best channel, I think is a really big important growth area for for remote work especially.
Yeah, I think, you know, that's a challenge with just about anybody especially, I mean, you know, when you have and a lot of instances with larger companies that have bigger employee basis, you end up with, you know, multiple generations at a given company and all of those different generations all kind of like, you know, communicate differently. I have a I have a 20 year old who works for me right now, and he does his does not answer the phone ever. Right? But he'll text you back in five seconds.
So funny make sense for me as well. And he works for us as well. That's That's ironic. But yeah, he's I mean, he's totally like, and now he won't even look at his, uh, his tax to look at like discord or like other other tools. Yeah, it's Yeah, you just got to have to learn where to where to? I mean, how to plug things in.
Yeah, it's so funny is this like some some of these younger kids, they literally live their lives with headphones in their ears constantly 24 hours a day, seven days a week. And, you know, they're their mode of communication is like, yo, DM me on Insta? I'm like, What? I'm like, I'm old. What did what the heck did you just say? I totally feel that even even my seven year old daughter comes home with words. And I'm like, What is that? Where did you hear that? Is that bad? Good? I don't even know what it is. The urban dictionary is your friend.
Exactly. No, it's funny. I mean, it's a probably similar age. It's like, having people work for us. I mean, that better the same age as my kids, too. So it's, it's funny to look at and be like, Okay, wait, I mean, it's like, can't be in dad mode now have to be in boss mode, even though they're the same age.
You know, I find that difficult every now and again, too, when I have one of my younger employees, and they have like a car problem, or, you know, they're complaining about the roommate situation or whatever. And I'm like, I could totally answer this question for you. But I don't really want to get in, you know, I don't want I don't want you to think it's because I've actually had an employee, I gave them some advice. And they got mad at me. Oh, why a younger employee? You know, don't don't tell me how to live my life. This is for me to figure out and I was like, Oh, I wasn't telling you what to do. I just was giving you a suggestion.
The whole other podcast series, it's like dad bosses or something.
We should probably start that with you. And I could pray co host that one? Man, well, it's been great talking with you, we got just a couple more minutes left in the program here. You know, tell people where they can find out information about you, the company, maybe your social media, that kind of stuff. And, you know, where, you know, people because a lot of people that listen to this show are technology people, they're entrepreneurs, you know, they may have a use for what you're doing. And so we know where can people find more information about Kirk the company and and, you know, follow along with the, you know, the growth of, of the company with unstruck. And what you guys are doing?
Yeah, for sure. No, it says a company website is it's just unstruck un, str Uk and.com. And so that's the company website, Kirk marble on LinkedIn, as well as Twitter. And then I think we're at unstruck on Twitter, as well. So that's our main stuff. But um, LinkedIn is probably the easiest way to get a hold of me. I'd love to I mean, I've met I mean a ton of people actually from doing podcasts and even from investors and customers. And I always take calls with anybody even I mean, just of talking about entrepreneurship and mean any any just advice I can give, always happy to, for my experience, but yeah, it's gonna be exciting. I'm really excited about our launch. I mean, we have a there's a conference, we're going to be going to in October, a reality capture network conference in Idaho that we're going to attend. And that's going to be our kind of launch party for the product. And then we
were at Idaho. Boise, all right, my, my family's from there from Idaho. Oh, man, all over the place. Yeah.
This one's gonna be cool. We actually we just hired somebody in Boise as well. And I'm right outside of Boise. But yeah, this one looks fun. It's gonna be all about 3d scanning and 3d, like, I know, comfort construction, and all that kind of stuff. I'm really excited about that conference. And that'll kind of be the first time that we're I mean, we're going to kind of push the button really widely, hopefully get a lot of signups after that. And then and then do a conference. It's data data day, Texas in Austin in January, which looks like a really good conferences. We're gonna attend and presented that one as well.
Yeah, that's awesome. You know, I got to go. Actually, I had the last event that I went to before. You know, COVID was in November of 2019. And I got to go to this really cool thing called Digital Hollywood. And I walked in, and we were broadcasting there, right? So we set up our broadcast table, and we're doing interviews with the vendors and different things. And I wasn't quite sure what digital Hollywood was right? I never been there before I walked into the door. And oh, man, I was in heaven. They had, they had augmented reality stuff all over the place, and, you know, all kinds of different things. Leveraging like, you know, the Unreal Engine for gaming. Yeah, yeah. And so how and how TV and film was leveraging it and how DreamWorks is using the Unreal Engine to create characters for cartoon based content and like, it was Really, really interesting. And then to see like a pod where you sit in there and you put on the goggles, and it gets you into an immersive destination, and it was like, it was the coolest experience I've ever had with. You know, any kind of, you know, virtual reality stuff because I'm sitting in this pod and you literally feel weightless. And the game was your Spider Man flying through downtown New York shooting your webs. And every time I come onto a building, the chair would move in the breeze would be so cool. But yeah. And so just you mentioned three dimensional stuff, and I couldn't help but remember the 3d rendering of man, what's a singer that orange hair that passed away like a year ago? Or maybe years ago?
We've figured it out. You mean?
I'm bad. I'm bad with it. The guy from the Prodigy? Yes, correct. Yeah. And in the lab and the labyrinth? Yeah. Okay. David Bowie. There we go. Oh, yeah. So they use the Unreal Engine to create a three dimensional character of David Bowie and put them in a video game? Oh, wow. Yeah, it was like super cool. And then, you know, the way that they do 3d imaging with green screens, and, you know, some of those kinds of things, too, which I thought was interesting. So you know, all that metadata is what you're taught, there's, there's not so much metadata that goes out of that to that. So I could probably see your stuff working in that space as well.
There's been, I mean, just even today, there was a I mean, a bunch of acquisitions that are happening in the 3d space, I'm closing the loop with like, the I mean, epic, and their engine. Yep, kind of even more business use cases for 3d. And I think you're gonna start to see a lot of kind of crossovers between business and entertainment, and a lot of ways and so yeah, I mean, we even I mean, the idea of as you're walking around your facility and looking at things like a conveyor belt, if we can show in augmented reality, what it looked like three months ago, and kind of overlay those things, and I think there's so much interesting stuff we can do in that realm, just to augment people's analysis of all this unstructured data.
Totally agree, Kirk. And man, what a great time I've had with you. I want to thank you for being on the show. Thanks for taking time out of your busy schedule. For sure. You know, I know I know, I know, being a startup entrepreneur, you're doing like 50,000 things at one time. So I appreciate your time and hanging out with us today, man. Thank you. Ladies and gentlemen, you've just listened to Kirk Marple founder and CEO at unstruck data, go check out their website at unstruck u n str uk.com. Find out what they got going on. I want to thank Kirk for joining us. And of course, our friends over at kick caster for hooking us up with this fantastic guest. And also I want to remind you guys, if you're listening on your favorite pod catcher, make sure you give us a nice review and give us five stars, not four because we're all five star human beings. And we appreciate everybody listening to the radio program. And don't forget to tune in next week right here on The Voice America variety channel. We'll be bringing it to you live at 12 o'clock Pacific 3pm. Eastern, thank you so much for tuning in. I'm Ryan treasure. And thanks for finding your frequency. Hey, what's up everybody? So glad you tuned into the show today. What a great show. It is. Like I said earlier in the show at the end, I was going to give you some more information on our live stereo session on the stereo app. Stereo app users can engage with the platform to listen in, seek out topics and join conversations about issues and ideas that interest them. There is no lack of content on that application. You can flip through many conversations, ask questions, join ones make your own wide ranging topics on stereo comedy pop culture, lifestyle sports business technology, the app can be downloaded for free by Apple and Android users. Once users download the app, they'll be able to create an avatar and a profile. Hi, it's so much fun making my avatar It was super cool. Users can submit the audio messages to hosts have conversations to join those conversations in real time. Finding your frequency we'll have a live audience interactive episode on stereo. We're going to be doing this every Friday at 2pm pacific time on the stereo app. Again, finding a frequency is going to be having a live audience interactive episode every week, Friday at 2pm. Pacific time we're gonna do question and answers and we're going to talk about technology. We're going to talk about business. We're talking about marketing. We're gonna talk about how people found that frequency in life and in business and why they decided to do what they do and take questions from people that are listening to the show and allow you guys to engage with us and I really hope to see you on stereo again. dario.com forward slash radio Ryan one live every Friday at 2pm pacific time. So again, come to stereo.com forward slash radio Ryan one. Once you get in there, follow me. Make sure you guys tune in to the show. Thanks for listening
