One thing that we learned in the pandemic is that remote work is here to stay, it is not going away. And if you're a small business owner, you might want to build a remote first business. There's a lot of advantages from costs, as well as the ability to reach talent that you wouldn't normally be able to reach just in your local geography. And today, we're talking about how you can build a strong remote first company, and unlocking unstructured data in your business as a key advantage. We sit down with Kirk maple, the founder of unstructured data, where he leads a remote first business has raised a ton of money over zoom and gets into some of the best practices when raising venture capital via zoom. And what unstructured data is and why it is so crucial to your business and give you the differentiator that leads to success. I'm William Glasser, CEO and co founder of ostrich and of course, your host of the Silicon Alley Podcast, where it is my job to talk to entrepreneurs, VCs and top performers to understand what it truly takes to grow and scale a business. Go ahead and pound that subscribe button so you get notified of new episodes air every Friday. Without further ado, I hope you enjoy this episode of the Silicon Alley podcast featuring v. Kirk maple. Are you interested in growing and scaling your business? Welcome to the Silicon Alley Podcast, where you'll hear from entrepreneurs and venture capitalists and top performers on what it truly takes to grow and scale a business. You'll walk away with actionable insights you can apply in your own business and life. Now, Dwayne glass, the CEO and co founder of ostrich and your host of the Silicon Alley podcast. Kirk, welcome to Silicon Alley podcast. Super excited to have you on today. Yes, I'm here looking forward to it. Yeah, and I'm really excited to sit down and talk to you for a couple of reasons. One, you've got a deep technical experience, which I think is is really, really interesting and brings a different flavor to the podcast and what we have had on recently in terms of entrepreneurs that tend to be more business focused. But you also are a founder and business focus in that regard been in technology leadership. So you know, I'm interested to dive in as someone that does have, you know, text, technical expertise, what has been the process like rebuilding unstruck data? And then, you know, obviously, let us know what the business is for the audience's context.
For sure, yeah, no unstructured data. It's what we're calling it as an unstructured data warehouse. So it's kind of like Google Photos on steroids for industries. So the way you can put all your media into a SaaS service, and everything from 3d to images to document. So it's it, the process was really interesting. So I came from the media and entertainment industry, like software industry, had a video transcoding company for a number of years and had worked at Microsoft and all that. But I started to see all these parallels in the kind of goat industry side of the world that was having a lot of the same pain points that we were solving for the broadcaster's in studios. And from working at General Motors working at a bunch of I was CTO and VP at a few different companies, I kept seeing this threat of everybody's trying to build it themselves. And that was really what I finally I started working on the side, kind of a side project for fun, and I still write code, like every day. It's It was kind of a passion project. And then it just was the right time and was able to spin this out, get funding for it, and you haven't worked cranking away for release in like about four to six weeks. So it's a it's been exciting journey.
Yeah, I know, that's exciting. And I love the ability to blend your background, both in media, and then getting into like, you know, manufacturing, industrial and industry side. And, you know, those are two worlds that don't really necessarily, you know, have a lot of overlaps. Well, you
think so. And that's what I was really surprised where, I mean, we had been building software for I mean, like NBC and ESPN and folks like that for years. And there was all these kind of needs around I mean, what they're kind of technical needs were and when I started to look at, okay, well, video coming off of a autonomous vehicle, there's actually I mean, a ton of volume there. But the data streams that are recorded are actually kind of similar to closed captioning, in a way where Closed captioning is kind of time synchronized text. And if you think about time synchronized data or vehicle of like, I mean, what we're heading it's on or, I mean temperature, or kind of classic IoT, things like that. I think that's where I kind of came in as a sort of naive person starting to see Wow, there's all these parallels there. But there isn't good tooling around it. And so that's where I really started to hit on of like, especially at GM even we were trying to build something like this and I've talked time and time again, even to somebody yesterday big company, well no company, building it in house. And so I just that's where I think there's a mean our thesis is there's a sort of picks and shovels kind of just build a platform and build then have companies build on top of it. model that we're super focused on right now.
That's awesome. Yeah, so giving the tools to these companies in order to build a great product so they you know, they don't have to go start from scratch. They've got a strong base. Flying and can utilize unstructured unstructured tools in order to do that. Yeah,
it's difficult having I mean, selling the platform is always very difficult. And so I've had that experience in the past where you have to have a UI have to have a front end app, a good application to get people into it. But the power is kind of behind the scenes. So we're doing both I mean, we have we have a really nice web application we're building that gives them access to exploring your data. But also it's there's a very powerful platform. But over time, we expect people to kind of go directly to the platform, because that's kind of what I saw in the past and be you have to build those.
Yeah, no, that makes sense. So the application layer is something that's easy, you know, get value right away, something that people can access, and then platform is folks that will actually build their own applications on top of
it. Yep. Yep. Exactly. Still just working with design partners. I mean, kind of, it's everybody from ports to chemical companies to manufacturing companies, we're talking to you. But they're dealing with masses of data, I mean, hundreds of terabytes of data that they just don't have organized. Well, I mean, everybody has great tooling for what what is the SQL database look like? Or what I mean, it's a glorified spreadsheet in a lot of ways. It's, I mean, when you're dealing with files, and all of these different types, I mean, even the 3d and things like that massive work just to even get access to the data. So yeah, no,
absolutely. And so you've kind of alluded to this and touched on it. So could you clearly define what you mean by unstructured data versus structured data?
Yeah, I mean, it's, it is a term that I mean, different people have different things. I mean, it's, you see unstructured data used a lot in like logs, and it's just kind of logs for software and things like that. But for us, it's you could sort of replace it with media files, per se, it's everything, from images, to video, to 3d to documents, typically, it's stored in cloud storage, like object storage, like an s3 bucket or a data lake. But what we see is I mean, it's, it is semi structured in a way that there is always metadata, I mean, you can get Exif tags out of your photos and know your GPS location. And so our first pass of what we do is bring that content in and try and grab whatever structured data we can find. And so we call that the entity extraction phase. And then we actually align everything into a knowledge graph. So we can create edges between common things, and we've kind of simplified it down to just a tagging model, basically, extract tags from anything, documents, images, whatever, and then try and correlate all those those pieces together.
Gotcha. Okay, so yeah, using the bits of information for, you know, unstructured data that that are structured, the meta tags, when things were created, file names, all the location data, all that sort of that sort of information, and then using that to then go deeper and get into the things that are harder when you're talking about like, closed captioning, for example, on video, and, you know, applying some of those principles to unstructured data.
And then it'd be machine learning, computer vision, all those things kind of come in as a second layer of, Okay, once you have that data, what's inside of it? And how can you extract more information from it, that's where things really get interesting. So you kind of have to have that meat potatoes of the basic data management. But after that, I mean, that's, that's where the sky's the limit on what we can actually do with that. So,
gosh, so when you're thinking about like, the value proposition to the business, then it's around the, obviously you can structure unstructured data, which is great. You can sort through it, but it's being able to make decisions, or what do you see is the real, like pain point or problem that you're solving?
That's great question. I mean, that's we're taking this in phases. So the first phase of the releases, okay, the the funnel to get data into the platform, and then it's the extraction of information, the or the auto organization of that data. But you're right, what do you do with that data? That's the next phase. And we're looking at that as two ways. One is collaboration, the ability to have team members comment on an image or document or set of data and share data, almost like a sort of a slack like collaboration model around their data. That's that's one thing. But also, the other thing we've heard from customers is notifications. We want to be told when something interesting happens, because they have this flood of data coming in. It's I mean, it's the Google Alerts model. It's got a ton of data coming in, how do I I mean, tell someone or alert someone when something interesting is happening. That's really the next phase.
Gotcha. That makes sense. So when you think about your target customers, so you mentioned the different industries that you serve. So who is at right now, whether it's with the application or the platform, at the moment, who's who's that target? use cases that developers are?
Yeah, it's so initially, and this is where it's tricky. So we're actually hitting out at line of business users initially. So as you can say, a visual inspection is kind of the the sort of broad use case that we're looking at if anybody inspecting things in the real world. So somebody want a maintenance engineer a asset, health and integrity person. As a property management, we're talking to a major university where they're actually pretty savvy. I mean, they're they've hand built a bunch of this kind of stuff with an off the shelf databases and like ArcGIS, or some kind of system like that. But it's, they're having to cobble together solutions to manage imagery, and then the data all related to that in that geospatial world. And that's what's really interesting to be like, if it could be University group, and it's super savvy is building something like this themselves. Like there has to be a need for better tool, because I mean, it's in this kind of low code, no code world, I mean, sure. I mean, any line of business user can build a build an app, but this just seems like something that should exist in the market. And I mean, anybody from from a commercial real estate or anything like that, drones are becoming very commonplace. So it's, it's really interesting to see, I mean, the numbers, the volumes of data being captured or just going up and up.
Yeah, no, absolutely. And it's interesting when you apply, you mentioned that some of those use cases around university or property managers being able to look at your videos, properties, images and and make decisions around maintenance or whatever, whatever those decisions. How does that work? Is it up to you got something like called graphical search, right? or thinking about like that model of searching data? Can you talk a little bit about that and expand?
Sure. It's interesting. So we we kind of have our North Star is kind of what we call the triad of time geospace like geospatial data, and metadata. And so all of everything is organized where the data comes in, and we index it against those three kind of axes. So you can explore your data over time, we have a nice histogram, you can zoom into and get down even like I mean, what was taken within this hour, or see like bar chart, and all those kind of nice graphs of over time, we have some good map interfaces. So you can see like cluster data on a map, you can create geo fences, like draw a polygon, like on Zillow, or something like that. And then very deep metadata search. So you can do I mean, show me anything taken on a drone, like by a drone, and we infer that, okay, this was taken using a drone, we can even get down to like the make, or the model of the camera, things like that, as well as full text search. So it's those three things coming together, you can then persist it as a view, we call it. And so it's kind of like a saved search customers initially can just, they just throw their data, they don't do anything else. And then they can explore it in those kind of three realms. And that's just the start. And then it's they can add their own tags, they can I mean, right now we have a bunch of out of the box ml and computer vision algorithms. So we do text analytics, we auto tag, but the goal is then to have the developers or at least ml data engineers plug in their own models later. And so documents are a great example, I was doing some testing last night on very kind of industrial, like inspection reports at a facility. And there's so much language in there that the normal NLP algorithms don't grab, they don't know what the terms are. And we've seen that like in the medical community and things like that, where there's people are building models for those verticals, that's an area we know we're gonna have to do more and be able to be more pluggable, because every vertical has its own language. And it's only useful if you can, if you can sort of focus focus the effort. So yeah, no,
absolutely that makes makes a lot of sense. And being able to combine document data, and then you try to try to navigate the NLP world with, with all the different terms and terminologies is definitely, definitely a challenge. Is this where the platform plays so microservices is is it sounds like as part of the part of the ability to, you know, build on the platform? Can you talk a little bit about what microservices are, and how that works for the platform?
Yeah, I mean, we are a, quote, serverless architecture. So there's no virtual machines, there's no actual servers that we're managing, it's all little chunks of functionality. And it's all event driven, as well. So everything is triggered by some other event. And so that's an intentional architectural decision that makes it very scalable. And so that we're running on Microsoft Azure. Today, we're actually Microsoft startups company. But we're built out in a way that, I mean, you could throw us 100,000 files, and it'll just burst and scale out. And so we'll start running on 1000 functions running in the background instead of 10. And then the nice thing about that is it collapses back to when it starts to go idle. So we have from a micro service standpoint, and we have different functional areas of the upload handler, the entity extraction, the entity enrichment, all of those are, I mean, really, they're almost the function level, very, very granular level, but it makes it very easy to plug things in. So the ability for us to plug in UML model and just kind of fit into that workflow graph very easily. And it's all asynchronous. So you can sort of fire things off. Things happen, they put new events on the queue, they get handled. And so it's, it's super scalable in that sense. I mean, there's the one downside of serverless is there's a little bit of cold start time. So as it bursts, you have to kind of wait until it senses that it needs to burst out. So there's some tuning always to be done around that. Yeah, I mean, we've had good success with with this architecture so far. And I think it's going to work for us in the in the long term.
Yeah, no, that's really interesting. And, yeah, like the ability to, you know, use various functions that you've got down to that level. So for full context, I'm not an engineer, but built a tech product using low code and no code tools. And like, you know, learn a little bit to be able to navigate, but the ability to get down to a granular level was part of the reason that I use the platform that I use, because there's a lot of no code, low code tools out there that are like, super, you can't do a lot of customization. But it sounds like you've got the ability to do some really deep customization work, but it's still like, I could hop on there and probably don't want to say, right,
yeah, I mean, that's the plan. I mean, there's an API, it's graph qL, it's pretty standard API structure that we're using. So our front end just uses graph QL. But when we're gonna offer that graph, qL API to customers as well. So you could build I mean, use any new code tool loco tool out there, I just have to retool this week is a really nice, I mean, kind of internal tooling. So if customers are kind of building internal apps and things like that, they have direct access to graph QL. And so they could just talk, you could build something really easily on our API. And I mean, it's and that's where we really want to get to is there's a lot of, quote, Shadow it going on at these big facilities where people are kind of cobbling together different things. And if we can be that kind of background platform, a lot of different teams can just kind of invent what that last 20% like what that last mile is. I mean, there's so much capability once you get the data in there structured, interesting way. We want to like enable an ecosystem around that. That's awesome.
Yeah, no, I love that. Kirk, in in terms of how you built the platform, can you talk about your journey of, of actually developing the product? Or the idea? Like you kind of sounded like you were building some of these things, one off for different companies. But can you talk about the that process? Because I'm really curious to hear how you've built up to this point?
Yeah, it's interesting. I mean, so I sold my company, my last company in 2012. And so that was a video transcoding company work for the buyers for like three years, and was just super focused on that. And then finally, it was just time to do something else. And so that was like 2015, I ended up starting a little LLC on the side actually worked on that like almost full time for almost a year in the cloud services space. So I had been doing all on premise software, like, it was all I mean, you'd run it in your data center, and all that kind of stuff. And so I wanted to get a lot of experience more building for the cloud. And so that's where I started to get a lot more agile experience and started to learn the patterns more in cloud services. And then just happened to do a great job at General Motors where they needed somebody with a video background, I wanted to learn new technologies, it was a great fit, we were essentially trying to build a lot of what is in this kind of unstructured thesis of I mean, unstructured data management, and we were dealing with the data off the cruise vehicles. But I was starting to see I mean, there's a lot of I mean, use cases across multiple industries. And so I ended up just getting recruited to a couple different jobs like a sports data analytics company, a drone analytics company. But in the background I was working my hobby was like, writing code for I was building a podcast discovery platform, and based on a knowledge graph, so it was essentially extracting, meaning the podcasts were the unstructured data, that I was extracting things from creating a knowledge graph, built an API. So I was actually building this product on the side just for fun. And the last, I guess, around COVID times, I started to realize number one, I was at a company where they were, we were trying to incubate a product kind of in this space. And so the COVID hit things that were kind of rotating around, I was trying to figure out what to do next. And realize that we can kind of pick the best thoughts that we're looking at in that drone space and that space, tie it in with the code I'd already written. And then was able to find seed funding for it. And so the back end of our product is our is code that I wrote over the last five years. So we actually came in with this big bucket of code, day one in like, whatever, February of this year. And that's why we're able to get to market really fast. I mean, we only closed our funding in March, but I had, I mean a couple 100,000 lines of code already written day one and now we've mostly been focusing on being very design led from the top down, hired a great design leader. I have a great engineering team. We've all worked together before. And so now it's all about just kind of finalizing everything and packaging it and just getting a really high quality product out there.
Gotcha. No, yeah, that's really interesting. So you've been working on this as a project for five years. Yeah, ended up ended up being able to utilize it. So what was that process like of raising seed funding? So you already you had code written. So you had something built that you that you could play with you had at least some other ideas of how you can apply it not only to podcasting, but to other industries? Talk about the process of raising that seed round.
Yeah, it was interesting times, I mean, so I was at a company drone image analytics company. And we were thinking of spinning off, I mean, this concept, and it was kind of one of those things of like, how do we go find funding for it? Do we split it up ourselves. And, I mean, it was all kind of right in the in the guts of COVID, at that time, and so it ended up being and probably started in around October, November of last year, and ended up starting to just get contacts of people that I mean, starting to show the concept to kind of broaden and refine the concept, and it did rotate around a little bit, but it's still pretty much in the same, I mean, the same direction that we're heading. And so much of fundraising is just knowing the right people, and just having some really great advisors and getting the right contacts and things start to lock down after the new year, and ended up having a great lead investor ABC, basically led the round. And then we found a bunch of other great investors essentially did like a two part seed round, and then ended up shell ventures, like oil and gas actually came in at the end of our of our seed round to finish it out. So I mean, we have some great strategic partners as well as angels. And then I mean, a great lead investor as well. So I mean, it's definitely a process, I think, I mean, I did all of the pitching on zoom, which is a weird thing to do. I'd done some fundraising back in the day, I mean, like, 10 plus years ago for my own company, but we had actually bootstrapped that whole company. So this is the first time actually raising money. And it was, I mean, it really it, I wouldn't say that it was that painful. Because I think people really got the idea. And they started to see the VCs or talk to her super savvy these days, in terms of understanding the data market, I mean, understanding how all the pieces fit together. I mean, it honestly wasn't as bad as I thought it was gonna be. Because people I mean, they really understood the need, and from from seeing what, what's out there today.
Now, that's awesome. Yeah, it's a pretty, pretty cool story, especially, you know, just how quickly your go from essentially just starting the company to seed round to, you know, launching, you know, end of July and sounds like you're able to get some really great partners, and what was that experience like, you know, fundraising via zoom for, for folks, because, you know, obviously, we're opening up now, but I don't think that the the zoom meetings for fundraising are not going to necessarily go away.
I mean, it's, it's an interesting thing, I mean, you can definitely pack a lot more in. I mean, that's, I think, both from the investor side, and from the founder side, it's a lot more efficient. We're a remote first company. So it's kind of right in our sweet spot of my old company was all remote first, this company is remote for so I mean, we're used to communicating this way. I think there's a lot of people out there that just aren't used to presenting themselves in in that that way. And, but also it kind of levels the playing field a bit. I mean, I think it actually makes people a bit more casual, you have really good conversations, there's not sort of this I mean, formality to it in a way of I mean, going to go into their offices and all that kind of stuff. It's everybody has a dog running around the background, like there's always like, I mean, this is good casualness that I think actually made the process more effective. And people were just kind of more thinking about the problems to solve than the formality of the sort of investor founder difference. Yeah, no,
I like that. I think I think you're right, it definitely does open up the playing field in terms of not having to go you know, have the funds to go spend to go travel to New York or Silicon Valley, and like, do that whole song and dance. And instead, you can connect with people all over the world.
I mean, one of our investors over Twitter and like, I mean, just, it's it's crazy how these different connections come in. But it's, it's people I never would have met normally, and was able to, to have a quick call with and I mean, even talk to people in Europe, I mean, talk to people in East Coast, and it's, yeah, it would be a much bigger process to try and do this. face to face. That's awesome.
Well, Kirk, in terms of the remote first, I'm curious to dig into that a little bit. So how, how does that work? How do you create and build a company and develop a team and kind of the culture in a remote first world?
I mean, it's it's definitely tricky, but I think I mean, tools, obviously are the big I mean, one of the first starting points of using slack using Google Docs very well. I think there's I mean, we've all worked together before and have met face to face face before. Not a ton. I mean, we we would come in for quarterly meetings and stuff like that. But I think it's, I mean, you people have to have good written written communication, to explain themselves well, because that's a lot of it. It's asynchronous communication to where I'm a late night person, and I have people in It was too early. And so like there's like, sometimes not overlap of like, if I'm working late at night, I leave a message for them, I have to be clear about that. And I have to like, I can't just sit and schedule a meeting to explain something to them, I have to be clear in writing, I think, honestly, that kind of communication is key, the attention to detail is very, very important. But also, I think we try and get together I mean, we do like an all hands meeting, where I mean everybody's video, like I have to do a video on, there's a balance between kind of that zoom fatigue thing where we do dailies on daily meetings on Discord. And so we can kind of have ways to chat that way we like since we're a tech comm software company, we use JIRA, and all those kind of tools to it's, it's really making sure you have the artifacts burden very well that you can hand off things to other people, very crisply. So it's, I mean, culture wise, I mean, I think it's, it's it's definitely a challenge. I mean, you have to not ever, it's not everybody's cup of tea, for sure. I know, there's a lot of people, even when I was at my old company, I was living in LA and commuting up to San Francisco, like every week for the first six months when I started. And it was a very quiet office. And I was so surprised. I was like, look at me, we have an office situation, but there's just no energy. Communication and slack was even planned. And it was funny, because comparing that to my remote first company that I've been happy for, we're chatting, I mean, we're all like sending and sending gifts to each other. And I mean joking around. And it's and it's just like, so much more energy, even in a remote environment. So it's, it's, I mean, it's kind of a culture you want to create has to kind of come out in that. And so we we try and be I mean, at least light hearted about it and not be like so serious all the time. And I think it helps kind of bring everybody into that equation. And and we've been good. I mean, we've brought on contractors, we've I mean, who then we promoted to full time. I mean, we have people all over the country right now people have to be aware that we are we are remote first. Yeah,
no, that makes sense. And great remote remote first isn't for everyone. And obviously not every job company can can do that, depending on what they do what they focus on. And as you said, the culture but it sounds like there are a lot of advantages. And I think we've we've learned that throughout your last 18 months, the ability to continue to, you know, get things done. And you know how you have to manage productivity a little bit differently, and who's suited for remote work first.
And then we are going to do and we are planning quarterly retreats. So we'll get everybody together quarterly. I mean, we were trying to figure out what our first one is going to be now. But I do think that is important. Like I mean, going 100% remote, like where you literally never meet anybody is a bit difficult. But even having a day or two just to see their face, even just understand people's personality better. Because I mean, some people can be a little dry, they can be a little sarcastic. We could be whatever I'm on slack and not having that face to the name. It's a little hard to parse the text sometimes and that face to face. Really I mean, it honestly can help a lot.
Yeah, absolutely. Absolutely. So yeah, combining that the remote remote first for the most part, but still having time where you can actually get to know people in person and understand personality and connect in a different way than you can do via the screen. So one of the things that you mentioned, Kirk, when you were talking about, you know, fundraising is that you have some angels, but you also mentioned some strategic investor. So you mentioned shell, can you talk about the difference of those types of investors and maybe the role that they're playing and the rollout of the platform?
Yeah, I mean, it's, it's interesting. So I met shell through my previous company, they've been investing them. I mean, what I've seen is they have a very good data science group. And there's some folks I met there that mean they're super innovative. I mean, they're, they're really pushing into alternative energy areas. And they're they're looking out, I mean, that, that multi year approach, and they have a very strong Innovation Group. So that is, I mean, they really are trying to keep their thumb on how to optimize their workflows. And so I actually did a talk at Shell, maybe a year and a half an hour ago. And they're taking videos of the undersea floor for their oil pipelines. And we had been helping them with some computer vision models and things like that. But it's really interesting, where I mean, this, the volume of data at a company like that is just insane. And so we had been, that was another data point where I saw I look, I mean, there's just files just, it's like straight off the floor. I mean, there's no organization to them. They're just like in a big bucket somewhere. And there's no rhyme or reason to say, hey, I want to go to this GPS location in this month, and show me all the videos tagged with a rock that we saw computer vision, like, if we could just do that. And like that compressing of all that exploration time of the data is huge for a company like and those are the kinds of things we're enabling in RV one is just the ability to throw a ton of data at us. We auto organize it and then the ability to visualize it and search it and then over time, what they want to do is build custom models to plug into This platform that's we're hearing a lot from the bigger companies. But yeah, I mean, I think it's for us the value of having a strategic partner like that is setting a long term product roadmap, kind of seeing what their blue sky is, and trying to align our long term roadmap. I mean, I'm kind of CEO and CTO right now. But I'm also kind of functioning as head of product. So I'm kind of doing, doing a long term vision. But it's really trying to just make sure those quarterly progressions are make sense. And we're not building things too early or too late. And that kind of thing. And, and I love I mean, this is where my business and tech kind of blends is actually love talking to customers, because it's the best way to do product planning. And that side of it is the more meetings we can have, I'm always getting ideas. I mean, it's, it's, you can hear them say something and I'm like, oh, wow, that fills in another little gap of something that I was thinking about, I wasn't sure if anybody would want. And so just having more people like that on our share, I mean, what do you call a cap table? Or I mean, just people we meet is super valuable. So
that's awesome. Yeah, no, and you have the ability to have someone that can help you figure out the product, cuz they're also going to be they're obviously an investor, but they're also invested as, as a user and a consumer of the product, which allows you to set that direction. And, again, they're gonna be thinking about long term vision as well, not just, we want this now, or why doesn't this work? It's like, how do we build this so that it can we can get to the, to where we see that the platform can be in the next, you know, three? 510 years?
Exactly. I mean, for us, it's a long term bet. I mean, this is a we're taking a pretty big swing, I mean, we I mean, have pretty high goals. For the company. I personally have high goals for the company. But I think there's there's such a need here. But it's also we kind of have to meet customers where they are in their lifecycle, because some people just want a better version of iPhoto. But for like, I mean dealing with 100,000 photos, not just 1000, some people want like this entire machine learning lifecycle application. And so we're trying to fit I mean, both those models, kind of as an easy button, and as a very professional developer focus tool. So
yeah, so how do you balance that right? Like, as you said, Your CTO CEO essentially had a product like how do you balance set goals, and manage both, you know, the business decisions, as well as the, you know, technology decisions that need to be made?
Yeah, I mean, it's kind of funny, I almost worked two jobs a day. So like, the first part of my day is kind of product and just business stuff, like dealing with paper, a little bit of paperwork, but also, there's so many better services out there for companies now to automate all that. I mean, we love gussto, and we love ramp and all these kinds of things that I mean, you can run your business on the SAS services, and it saves so much time. I mean, so I'm so thankful to have like those kind of products that I could just like delegate mundane stuff that I would probably procrastinate on. But then in terms of what that leaves me with is, I spend a good chunk of my day just talking about like reviewing the current state of the product, current roadmap, but also I'm doing a ton of research about the market and doing a ton of reading, listening to podcasts, trying to anticipate where the kind of where the pucks going, and that sense, but it is a balance. And so we really try and be very iterative on it, where we have a quarterly roadmap, we broken it down into like two weeks sprints. Now, we reorganize, I mean a little bit like we we really talk it through I mean, we have meetings every week of Okay, is this still the order that we think and as we get new data points from customers, we haven't changed too much. But it's kind of that negotiation of what the MVP is, like, what is the line, and I've been pretty aggressive about like what I think we need to ship with, but I'd rather be like, shoot a little too far, and then have to bring it back. And we just negotiate through that stuff with with the other leads. But there's what I mean, you're always learning and you're always coming up with something of like, like we heard, like single tenant like actually deploying in a customer's cloud tenant has come up, we've heard it a couple times, we don't know if it's like a must have, or like a should have. And it's always right. Those are the things on the bubble that we're always trying to, like, you want to get ahead of it enough, because you can't wait too late. Like we can't wait to make that decision to support that until December if somebody needs to January. So that's a lot of my thought process is just I mean refinement refinement, of when do we need to bring in the things early and not do work with throw away? I mean, okay, it's not easy, but it's just becomes, as long as you're, you're kind of thinking about it all the time, then you can react pretty quickly.
Yeah, no, that makes sense. And yeah, me being able to prioritize and try to take little data points from here from there something that you hear from a customer and try to verify, okay, is this just their problem? Or is this something that everyone is going to, you know, going to need, and we actually do need to build this versus just this one customer that has some weird, unique use case.
And it's weird. I mean, we're, I mean, we're cloud hosted. I mean, we're on Azure. Today we have plans to go multi cloud and deploy on AWS or GCP. We can take data from those platforms. But our code is not running on AWS today. And we've heard it from a couple customers, I'm like, No, well we need to do is shop. I mean, our security people would want to solve me running there. And we don't know yet. Is that one in a million customers? Is it one in five customers that want that? And so we're still trying to refine that that process. But also, we have had a little bit of pre planning to say, Well, if we would need to do this, what does it mean? And so we have kind of brainstorming meetings, we just call them where, let's just try and get ahead of a couple of these things. And we don't know what, like when we're going to implement it. But we've kind of pre chewed the work, I would say, so. Yeah, that's interesting.
I really, I really liked that. I don't know that I've heard too many people talk about, you know, thinking through different scenarios to having a brainstorming session and saying, okay, we know what we would do if this does become something that we need to do. And let's shelve it for now. That's
interesting. To me. It's tricky. I think it's, you kind of have to be accepting of that, where I try not to give the team work. That is just throwaway work. And, but there may be things that maybe we need to do a little bit of this now we know we're going to come back to it later. And I mean, we've even done things like pre mortems, where, okay, we got this product, like what are the failure points are thinking through like, I mean, we're may we've screwed up, I mean, where maybe the problems and so yeah, I mean, we've been having some really interesting discussions, trying to be a little more forth. I mean, forthright are ahead of the game in some of these areas, but I mean, we're still small, I think we're like a full timers right now. But just trying to I mean, keep everybody's focused on the today, but also kind of what is the next thing in front of us?
Make sense? Makes a lot of sense, Kirk. So thinking about the future, you know, what is? What does success look like? How do you define success?
No, it's a great question. Um, it's, I mean, for me, it's really about market acceptance. It's, I love having customers use the products that I build. And so I mean, that pays off, obviously, financially, and all those kind of things later, but it really starts with, you got to have customers using your product, none of the other things make sense, like you're not going to get more funding, you're not going to get an exit. And so I really focus on if customers are happy, using the application, we're good seems like a pleasant application to use. That's one part of success. And that's why I've I mean, I treat us as very design lead, like UX heavy UI heavy, like we're thinking through that functionality, because that's what's going to drive traction. But also, it's just becoming a de facto standard, like, like a snowflake or something like that, or like a data bricks. Like, I mean, if you're talking blue sky success, I mean, for us, that's where I want to be, is kind of one of the matte like, probably like primary tier, data engineering and data product providers. And I mean, there's a lot to do once you get there. But it's, it's really about traction. And from a confidence level, I think I'm pretty pragmatic about it. But I think I've heard enough data points that I know there's a problem here. I think we just got to carve it down to the right solvable problem for the right, how much what's the cost for the customer, and all that kind of stuff. But it's like with the investment with all the customer discussions we've had, there's definitely an overlooked area here. Like you either have to do it yourself. That's what we always say, like working competing against Python programmers, in a sense. Typically, I mean, the way you're going to solve this problem is to write code. And or you're going to kind of cobble together some tools that don't scale. And so I mean, but there's also a ton of risk. I mean, one of the big boys could come out and try and stumble over this area. But I believe that I have a different perspective coming from the media entertainment world, like I'm not a data scientist, like I'm not, there's a lot of very data science, first products that are coming out, I'm almost naively coming at it from an opposite direction, where I'm thinking media First, the files first. And I think that gives us that gives us a bit of a different perspective.
Yeah, I appreciate you sharing that. And obviously, you know, vision of being that standard would be would be amazing. But even boiling it down to just customers using the product is the first first most important thing really interesting and then kind of that differentiator that you said of being focused on the media first and the files even how do you view this problem this space differently than everyone else is doing and I think that that's so critical when you get into a heavily heavily trafficked or heavily sought after area where there's just a lot of a lot of noise in the data space.
And it mean I sold a lot in the broadcast world and things like that. It's mean so much of how much they could process was based on good storage, good networking, like it was almost more of a solutions engineering problem than a software problem. And now so much that's commoditized these days, I mean, things I like sort of quote invented back in the day in terms of like, splitting up files into little chunks and putting them back together. Like that's how Netflix includes videos today. Like it's all that's become commodity and but back 10 years ago that was unique, and so having to come up with something that still has value and isn't just going to get stomped on. By I mean every other vendor we just needed. And we talked about being opinionated, a lot. Like there's some level of what we're doing that we're very opinionated about. And we're not going to change depending on what feedback we get. Because we really believe that this is I mean, people might not see it. But this is really the what people want should be doing. But then there's the other percentage of it of taking in that feedback and adjusting and making sure that oh, maybe we screwed up, you got to balance that. But I do think you have to be opinionated about some part of it. Otherwise, you just get dragged around by every customer meeting you have.
Yeah, no, that makes a lot of sense. Yeah, having that opinion. And here's what we believe. And it allows people to say, well, we believe that too. And yeah, that makes sense, or it doesn't. And I think one of the other things that you mentioned earlier on in the conversation was that you're focusing on line of business, right as being the customer where it sounds like a lot of folks in this space are more focused on the data scientist or the that community versus how can a line of business create the applications that they need without having to go bring in resources internally to or externally to build the tools that they need.
Because we're I mean, basically that since we're focusing on the capture side, so we're going upstream, and I said, from day one of this product of, I mean, follow the data. And so we're actually building out a mobile app for basically streamlining data capture. So capturing images, capturing video, I mean, you get your geo tagging of the images, but we can do, you can put in tags, you can put in comments, you can. And then almost like a as you're walking around, you can use almost like a Foursquare kind of model, you can see what other media was captured around you. So like somebody at a port can walk around and see the data captured last week, or the last month, automatically. And then we want to encourage people to put in data through that method, it helps the fidelity of their data downstream. And so we're really just focused on Okay, this first kind of, say, half of the year, or the, I guess, the next six months, really, how do we get more data put into our warehouse? And then then really, collaboration on that data is key. But then the next part is what do you do with that data? Like you said before? Have you run more algorithms on it? I mean, do you notify when things happen, and there's just so many other interesting things we can do for alerting and things like that. And that's where people want to go, that's what customers want to go, we have to have the data first. And that's, that's really what we're starting with.
Yeah, no, that makes sense. That makes sense. Yeah, I can even see application where you're pulling in data that's not even related to someone putting it in themselves, right? publicly available stuff going in Google Maps, or whatever it is, someone posted on Twitter a video in Times Square, and here's, you know, pulling in all that other information, you can really make some very interesting decisions.
That was my whole thing of the podcast discovery platform where when live events, were saying, like I was a big, went to see a lot of like music and things like that. And I always thought, look, there's so much other information that's out there that I mean, you can't see what like it's, it's a concept of a graph, you can't see like, whatever. I mean, this band played at this, this event at this club, who played at that club before, and almost like a better bands in town kind of concept, to explore that graph of data that's being generated by all those live music events, and those bands and their collaborations. And, I mean, obviously, Spotify students, some of those bands and answering some of those, but I had a view of luck. There's more data there that if you could go scrape webpages and scrape Twitter like you're saying, and federate all this data, enrich all this data together. And honestly, that's the technique that now we're using for this industrial data. And the goal is, Oh, you want to go and rich with an API on their side for their equipment list or something like that, you can do that. So but the platform is rather generic. That I mean, maybe someday we'll come around and build a music app. I doubt it. But it's I mean, the guts of the product, honestly, are pretty can be used for any domain, but we're targeting this domain first.
Makes sense? That makes sense. So I want to transition a little bit here. So my, my focus outside of this podcast is on personal finance. And I always love to talk to entrepreneurs and get in kind of get their, their kind of take on on this. So I'd love to understand how you would describe your relationship with money
up and down. So I lived through the stock market crash of 2000. So I mean, I've definitely been on a higher level over the years, depending on I mean, where I've worked in stock market and stuff like that. I mean, I mean, my kids are a bit older now. I mean, it's in their early 20s now and so it's trying to pass along like my failure stuff, like, I mean, living within my means, or just I mean being smart about saving and stuff like that, I think. I mean, I didn't have a lot of good sort of family guidance, I guess I could say in personal finance, and that's one thing I think is so important these days is to teach that guidance and there's a lot of good tools out there now, for tracking that. I've definitely been having bootstrapped a company and we used to get be down to our last dollar and then get 50 grand the next day like it was just crazy up and down. Life. Have bootstrapping. And so I mean, I've kind of lived at all that since I kind of the one good thing I think is I kind of bring that frugality now to the company where we have a decent amount of money in the bank from our fundraise, but it's really treating it more like a bootstrap, like, I mean, you don't have to just blow money for money's sake. So I think I'm at this kind of stage of the game, I'm really looking at it as like, Okay, how long can we stretch this, but I'm really looking at every dollar of as an investment. And on my side, it's really just a mean, thankfully, my kids are all past college now. And I don't have to pay for that anymore. So it's up to you to kind of predict what the next the next several years ago.
Yeah, no, I appreciate you sharing. Yeah. And I think I think you're not alone in terms of not not being taught whether family school, having that kind of knowledge. It's something that unfortunately, we either have someone in your family who really is focused on it and knows it or, or you're kind of, you know, figure it out on your own. So it's, it's awesome that you've, you've had that and be able to learn from your experiences over the years and bootstrapped company and being able to apply that to your current situation.
Now, it's a I mean, failure is kind of the best teacher so I think, I mean, but it always makes the, the winds better, in a sense. And so I think it's at some point, you just, like, have to take those learnings. And I mean, especially when you have kids, it's like, make sure that they understand like, okay, here's where I screwed up four times. And I try not to do that again. So
yeah, makes sense. Makes sense? Yeah, at least pass those on. What would you say is the best investment when you're talking about Windows just a second ago? What would you say the best investment that you've made?
I mean, it's a time I mean, honestly, it's it's investing in the company. I mean, myself, investing myself, I guess is a better way to say it is, after I sold my company, I knew I wanted to do something else. But I knew I had to learning and I kind of knew there was different things I wanted to do to kind of build up to starting another company someday, I think just from a financial investment. I can't say I mean, I haven't been in the stock market vaccines for a while now just since the last crash, but it's the investing in yourself for the future. And making sure really, I mean, to me, time is more of a currency than money is this point. So it was making sure that spending time with your family properly and spending time on the business properly? And as an entrepreneur, that's always a hard lesson to learn as well. But I think I mean, I think it's now it's just trying to put your eggs in the right basket, in terms of of time or the money.
Yeah, I love that. Yeah. And it's what you're what you're doing with the assets that you have an unstructured data, right is do I miss a big component of that, and being able to look at that, but investing in yourself makes makes a lot of sense. Now, we don't always make good decisions, what would you say is the dumbest money mistake that you've made,
letting other people manage your money? When I was at Microsoft, Microsoft is doing really well. It just felt like free money. It was like Everybody is waiting for it to double again. And, and this is right before the crash in 2000. And I left Microsoft and got a recommendation for money manager and was like, Oh, cool. I mean, I'll He's like, diversify, do this, do that do this and thinking that someone else has your best interest. And they're not just trying to pump and dump you to, I mean, make their commission that was completely the dumbest thing I've ever done. And I ended up losing like most of it just from when it crashed and, and be it in a bunch of random other things if he had suggested and I just took my eye off the ball. And that's I mean, that's probably the one of the more painful lessons of with money, especially and this is something I wish I had gotten taught is just make sure you're not delegated responsibility. And even if you make stupid mistakes, at least make it make sure it's your mistake. Not that you just like allowed someone else to kind of like make a stupid mistake for us. Yeah,
no, I like that. Yeah, then you can at least learn and hopefully that's if you do make a mistake of that, and you're making it you can at least learn from it. And yeah, and make better decisions in the future.
Yeah, for sure. That was, and even just with the stock market stuff, I mean, just getting strung out on credit, where they push you to go over your bad, like, bounce a little bit for what you really want. And just being strong about setting a line and being comfortable with it.
Makes sense? Okay, this has been been a lot of fun. I really appreciate you sitting down and I want to leave you with the last word. So anything that you want to share with the audience, and then please let everyone know how they can connect with you outside of this podcast. Yeah, no,
it's been great. Really appreciate the conversation. And so the company unstruck data is, we're on Twitter, I think it's at unstruck. And then myself at Kirk Marple on Twitter and LinkedIn is a great way to get a hold of me as well.
Awesome. Well, thanks again for sitting down. This is a lot of fun. We covered a lot of ground in a short amount of time and I appreciate you, you know, diving in deep on what you're building.
Thank you so much. I
appreciate it. On your way out. Please share the podcast with others is the only way that the community grows and others hear these incredible stories from entrepreneurs and top performers. And of course, pound that subscribe button so you're notified when new episodes drop every Friday. I'm William Glasser, CEO and co founder of ostrich and of course your hosts of the Silicon Alley podcast. Have a very profitable day you got on top too.
But still you as it's a calling in a circle saying I'll never leave this place somewhere it's good to serve on the right side over and over.