02:30 Hey, welcome back to the show. And today we've got someone coming all the way from the beautiful PNW. That's the Pacific Northwest. Kirk, tell us who you are and what do you do?

02:40 Yeah, thanks for having me here. I mean, Kirk Marple, founder and CEO of a company called Graphlit and I've been working in this space and in the Pacific Northwest after being at Microsoft.

02:52 Love it. So you're a transplant for Microsoft and you've evolved.

02:56 I have, I have. So definitely kind of cut my teeth there and I've been doing startup stuff for the last 20 years.

03:02 I love it. I love it. Tell us about that. What, what, what are you doing now? What is Graphlit? I mean, that can mean a lot of things. Talk to us. Who are you helping with this business?

03:12 So we're really helping developers. So we had gotten our start focused on what's called unstructured data. So anything from documents to web pages to audio and video. And we were building tools for cataloging that data. So essentially searching, visualizing it. And then LLMs came out, ChatGPT, and we realized how the kind of peanut butter and jelly go together. And so we're building this platform for developers to make it easier to use.

03:39 I love it, I love it. Everyone likes a PBJ and when you find, you know, honestly, I feel bad for people even 150 years ago that didn't know that peanut butter and chocolate could go together the way they do. You know what I mean? 

03:50 It's true, it's true.

03:51 It's kind of, it's an unfair thing. So you discovered that AI helps you to manipulate this data better and organize it in a better way and probably call it up better. Is that what you're talking about?

04:04 Exactly. I mean, it's so much of what you do with that data is just how do you find it? And we had started talking to people that had 20 years of data, just sitting on hard drives in their business. And these, it was in different formats and it was different. I mean, people, some people didn't even know what was there. And so we started focusing on just how do we kind of be the janitors for that data and try and figure out, okay, how do I even look at what's there? And so it's kind of like, We all do this on our computers and understand, okay, what files do I have? Where are they? And we were just doing this on a much, much bigger scale. And then using AI in the early days for some of this, but now it's even become more part of it.

04:42 I would imagine. I would imagine. I mean, correct me if I'm wrong, but I'm thinking like lawyers with lots of data for a case or doctors with notes of each client. Am I on the right path here or are we talking something different?

04:56 Yeah, I mean, because we want to stay very horizontal. And so we're not, what we say is we can handle that kind of middleware layer that just give people a good tool, a good API to build on and let those developers build their vertical businesses. So what we want to do is have developers kind of sitting their applications on top of us and then targeting healthcare, legal. And so we're not getting into those specific domains, but we can handle.

05:26 Middleware's a good place to be in most cases. I think it's a smart play.

05:30 Now, it's, and we had started building an application and then started realizing the value is really let anybody build that on the guts of what we built. And so that's what we've been about for the last 18 months. And the first 18 months were more like top level, top down. And then the last 18 had been more.

05:49 Nice, nice, I love it. This is really fun. I mean, everybody's obviously excited about AI, but when you consider all the data sets that people do sit on for years, because it's just so hard to organize, this is very cool. So it's able to take data that is, you said audio, video, I'm assuming you.

06:12 Images. I mean, and that's what I mean, we wouldn't be agnostic to any media format. And so my previous company was in the broadcast video space. And so we were dealing with like, I mean, video for ESPN a lot of closed captioning and audio. So this is like an extrapolation of what I've done for the previous kind of 10, 12 years. And it's really, I mean, it's cool because you can use computer vision to identify things and images. But now with all these vision models that are coming out with, it's not just text, it's what can I look at in a webpage screenshot and learn information.

06:52 Okay. So, this is awesome because you're solving a huge problem that most industries are dealing with and how do they take analog data, how do they take some of this older data. I mean, who'd have thought even 10 years ago that we'd be able to go back on video data and actually catalog that somehow. That's cool.

07:03 Yeah. Well, and the media management had been a thing for years. I mean, all the broadcasters, they have to have media archives. They have to be able to search it. And but a lot of it was at the kind of thin level of the file name or maybe a couple of labels kind of you might have. But what we've now seen and there's some companies in this space that are focused for broadcasters to like look at what the cast of characters are or, I mean, get into the media. And we've kind of been doing a similar thing, but more just developers. I mean, let people build cool applications on this on this data.

07:36 I love it.I love it. So your avatar as a business founder is you're going after the dev guys who are building those cool use cases, but making it easier for them to do so by plugging into how you manipulate the data.

07:49  Exactly. I mean, it's no different than, I mean, like people using a database. I mean, in the old days, like you'd have to write more to the metal and think about all the schemas and all that kind of stuff. And nowadays, I mean, it's just like much easier to use. And so we're kind of trying to do the same thing for unstructured data and just make it, make it a real commonplace thing.

08:11 Now I gotta ask, I mean, you did kind of elude that in the beginning, your thought was we're writing a solution, start to finish. Looking back on that decision, what triggered the decision first of all? And then, has that been a good thing for you? What do you think?

08:27 It's interesting to say it was always a hedge bet. I mean, I had actually started working on the platform prior to starting the company. And I actually wanted to build the platform and basically launched the platform initially, but it didn't feel like the market was ready for it. I mean, three years ago, people weren't talking about unstructured data. I had been CTO of a couple of different places building some tools like this. And it really got to a point where, okay, what could people absorb? So we thought, okay, this kind of iPhoto for industry concept, a lightweight cataloging tool. And I think it's, I mean, it was a cool product, but I think it was just a little bit early, or at least maybe it was, we were just selling it to the wrong people. Because the folks that could use it were still like SharePoint customers. And that's all they could, they just couldn't absorb the next layer of what we were trying to do. Yeah. And so what we started to see was, I mean, there was the innovators at these organizations were like, my God, this is amazing. But just the IT groups were like, well, we already paid three bucks a head for SharePoint. I don't know why we'd want to do this. So it was really just going back to my roots. This was always the idea of opening up the API was always there. But ended up doing a much more formal pivot where we deprecated the UI. We just took some time, maybe three or four months and just created a developer portal, you can get your own API key. And that's been the last year.

10:04 I love it. I honestly feel like that's such a smart move. And for those of you listening, it's okay that you could be a year, year and a half into your business development and actually pivot like this. This is a really, really smart play. And for a lot of reasons, I think it's a smart play. But for you, be thinking, okay, am I like her? Do I need to be thinking about this original context that you really enjoyed, which is I, I'm a wholeheartedly, I love when people open up an API that I can just kind of tag into and not to rewrite everything, but I can tag into you for what I need you for, pull the data back out and then do it. You know what I mean?

10:44 And we've really been hearing great feedback where, I mean, we keep saying we're competing against DIY because it's not that, I mean, there's parts of this you can do yourself, but you have to manage everything. It's like building a car from parts. I would say it's like go to Ikea and I mean, have to build the table. But for us, we're more just, hey, have the table delivered. I mean, have, it just, it works, it scales. And you also get the feature set that we can now offer to everyone and you get that volume change where we can offer new models and new capabilities to every one of our customers. And it just saves them so much time.

11:25 And are you a storage facility or are you just a process and stamp and tag and okay?

11:33 We're both. Yep, we're both. And so we can act essentially like a media archive, like a data archive. Sit on top of cloud storage. I mean, we have the compute layer. We are SaaS today. I mean, so it's just self -service, come in, upload your data, use the data. We are looking at doing more private cloud deployments next quarter, because that's what I've been in. I mean, for bigger customers, especially in the more sensitive areas, like healthcare, they want full control over this. And we always kind of expect that.

12:04 I was gonna say, I was gonna say PCI compliance and HIPAA and all those things are probably, yeah, that's kind of a place that gets a little touchy. A lot of people.

12:13 Exactly. So that's what we're we're in Azure shop and so we're going to be leveraging the Azure marketplace and really taking advantage of a lot of stuff there to sort of just run an instance of graphlet in their Azure subscription.

12:26 captainscouncil.com

14:06 I love it, I love it. This is very cool, I love the model, I love what you're doing with it. You know, as you, I looked at your LinkedIn profile, and for those that haven't, check it out, interesting background. You know, as you've been CTO of multiple companies, what was the transition like going from that role into being a CEO and, you know, Chief Technical Officer in your own organization. Is it pretty different? I mean, talk to us about that transition.

14:35 Yeah, I mean, hold on, let me just see, my video got weird for a second. There we go, the autofocus was out. You can clip that up. I mean, I think it's really interesting. I mean, being a smaller company founder, I think is not that different than being like a development manager or like a director or VP at a bigger company in some parts. I mean, in the product and engineering part, it's still a lot of the same day -to -day stuff, but then you also own all the resource management, all the finance side, all of that. So it's like one job plus two more jobs. But I actually love that part because I think the best product people see the customer side and have a better feel for what do customers really need. Pricing, sales, marketing. And I think that's the biggest thing that I've learned of when I was CTO or even just a manager in years past your head is down. I mean, you're focused on just one thing or a couple things. But now I get to see everything, which is even more exciting.

15:34  I love it, I love it. That's very cool stuff. I honestly, I love to ask that question because most of the people listening to this are similar. They have been an operational director of some sort within an organization and oftentimes the feedback I hear is similar to what you just described. But some people really don't like the customer facing part. I'm just a dev guy, I don't really want to talk to people but I want to build a cool product. I guess this being a community now that you're building between developers, it kind of keeps you close to home, does it not?

16:08 Yeah, I mean, I'm firsthand on support. I mean, dealing with customers, I mean, on a day -to -day basis, but it also gives product ideas. I mean, just somebody asks like, hey, okay, that's great. You do the X, Y, and Z, but can you do ABC? And often it's something maybe we thought of, we talked about six months ago, and now somebody's like, I really want that. And so I think it's, I love those conversations because I now get firsthand feedback. Or if it's something wrong, I mean, which hopefully isn't very often. I can, I mean, know really, okay, here's the pain they're facing. I mean, we had some things where like, okay, we were thinking our conversion rate from our free to our paid plans was not what we wanted. And then we started realizing it's just more sample apps. It wasn't even the product itself. It's people were just couldn't figure out how to get from A to B from the document, reading the docs to using it. And so we spent some time just building sample apps and putting up more open source sample code and our conversion rate like bounced up and it was just, and you have to see that from firsthand to really understand them.

17:17 Well, it's interesting because I was just going to ask you, what were some of the challenges that you've had in trying to grow that model? And it sounds like you just described one. I mean, it's obviously conversion from a free sample into a paid anything, SaaS or service or anything. You really have to dial into, okay, how do I keep the traction going once they've tried it? And how do we give them a good first experience? And so describe. Like first of all, how did you recognize that was even a problem? And second of all, what did that process look like? I mean, were you talking to clients? Were you asking the questions? What did you do to solve that problem?

17:56 I think the good thing is since I've been a developer so long, I mean, I know what products are easy to get to use and what, I mean, if I can read the docs, if I can start playing with them and I know I can get an API key quickly, I know where the friction is. And so I wanted to emulate the kind of best in class folks that have really good developer portals, really good documentation. People like Stripe. I mean, Stripe has incredible developer experience in APIs. And even people like SuperBase or folks like that, Vercell, I mean, people really, the community, I mean, a lot of them just really love them because they're easy to use. And so we basically just kind of stole ideas from all the best people and said, look, here's where we start. And we knew we wanted to be self -service. I mean, we knew we had to, I mean, I spent probably just an entire week just writing documentation just to get the first cut, but also trying to automate as much as I can too, is that's really where your time gets kind of stuck. But no, it's an end -to -end process. I mean, you really have to understand, try and put yourself in the head of the developer. And I think that's one of the things you can't do in a vacuum. Like people are, I mean, you may not understand why they think something's tricky or not understanding, but you want to try and fix it.

19:19 Right, right. Well, I can appreciate that. And those of you listening, if you're not constantly iterating based off customer feedback, chances are you're solving a problem that only you think is the problem. And more often than not, we as the developer or as a product designer or as the CEO, we oftentimes don't see exactly what our clients are dealing with from their perspective. We think we do, but we don't always do that. That loop of communication, how are you maintaining that?19:49 It's interesting because I had some ideas around this and I saw some stuff on Twitter. And one of the things that I find really valuable is like every error that we get goes into a Slack channel. And so we're able to just track things like that of every sign up, every everybody creating a project, even everybody deleting a project. So like if we get churned for some reason and somebody leaves, I think that trying to filter out kind of a signal channel. And it's funny, I kind of thought of it myself and then I saw people to be like, this is a really good idea. And I was like, it is a great idea. And it's, I mean, I've been, I mean, I'm able to see like, okay, overnight, I like if there's any exception, if the customer saw it or they didn't even see it, I can just quickly look at that to see kind of a sanitized summary of it and then go look in the logs. And that lets us get ahead of things so much where if there's some problem,And sometimes it's, yeah, like I said, customers may not even notice it, but I want to get ahead of it, fix it, make sure it's in the next room.

20:58 I love it, I love it. So I'm gonna spare everyone the details of your release model because I think that gets pretty nerdy, but I can appreciate it, I can appreciate it. But tell us about that, you've talked about the feedback loop, you've talked about adapting and making the changes. As you did go to market with this new model and things like that, were there unexpected challenges that kind of threw you off that you weren't really expecting?

21:25 I mean, I think I could, there's sort of two layers to that. I think at a higher level, I think just, I mean, even just from a release cycle, like how often do we release? How often do customers want us to release? It's such an interesting question. And we had to dial that in, where initially I was like, okay, I mean, we can release weekly if we wanted to, or even more often than that. But from a customer standpoint, I mean, a lot of times they don't want that turn. Like they don't want to have to adapt that if they just wrote their code, It always adds a little bit of risk. They may want to check a new release. So we kind of dial that in now to like twice a month. And we have a change log they can look at. I mean, we do our best to make no backward compatible problems. But it is that kind of balance of like how, I mean, you want to be responsive, but almost if you're too responsive, then that puts the weight of validating those on the customer is one area. But then there's other, I mean, one more sort of technical one is like with Microsoft Azure. Like there was some limitations I didn't realize at first. And so you go in and think, well, this is just going to scale. And, but then you realize you could only have X of this number of resources in this region. And I hadn't realized that upfront, like storage accounts, like you can only have so many per region. And we basically just had to do like around Robin, where we just kind of now spread them around different regions. But, it was fine for the first couple of months. But then once you start being like, wow, we have enough customers that we start hitting these limits and you get an error and you're like, my God, like we've now scaled beyond these quota limits. And I was able to get them to bump it up a couple of times. But at some point I'm like, okay, I got to bite the bullet and come up with a real solution here. But I mean, I love using managed services like that because it's so much we don't have to do.

23:23 Right, right. And that's really, I would imagine that part of your experience in dealing with a vendor like that has helped you kind of develop your own internal things that you want to be for your client. Is that wrong?

23:33 100%, 100%. I think that's a really good point where, I mean, I see this as, I mean, it's actually interesting. With my old company, I wanted our interface and our application to feel like a product from Apple or Avid or Adobe, like for in that space where the UX is tight, it looks really clean, crisp, and it's here, I'm emulating like an Azure or AWS where I'm like. Okay, this should be just something like if it was branded as one of the cloud vendors, you shouldn't be surprised. Like it should just, I mean, it should feel that quality. That's what we shoot for. And obviously they have a million more people than we do, but you still want to hone in on that.

24:21 Yeah, yeah, yeah. Now I was gonna say, for those of you listening who are really still in that development stage and you're trying to figure out the UI and trying to figure out that whole customer experience part of it, don't hesitate to do what Kirk just described and copying people that are doing it right. I mean, you all have your own customer experience as a customer of somebody else, right? If you like the way somebody's treating you, it's a good chance you should probably try to implement that in your business model. Have you done that, Kurt? Did I hit that all right?

24:49 No, yeah, no, that's it's perfectly true. And I think that's where I mean, when we first started building our developer portal, I looked at a couple of them and I sort of knew what we needed. You need to have sign up, create an account, get an API key, be able to delete accounts, those kind of simple things. And we basically did almost like a like a mood board in a way, like just screenshotting a bunch of websites, coming up with the pieces that I like, these I don't like. And at that time, we didn't even have a designer involved. So a lot of it was just make it look like that, which I know is painful for our dads. But then we were able to find a great designer who was actually in Europe and was able to just come on, refine our UI. And we just shipped that, I guess, a month or so ago. And it's been great. I mean, you have to be cost effective. I'm always trying to put the money in the right places at the right time, which is a whole other conversation, line of conversation.

25:48 I totally hear you and I'm curious, three years into this, would you have thought three years ago you'd go to Europe to get your designer for your software, for your middleware?

26:00 It's, I mean, I've dealt, I mean, or utilized, I mean, international developers for years. I mean, we had had folks in India, in Ukraine, and South America. But I think from a design aesthetic standpoint, I think that European developers are actually, or European designers are actually unique. And I see like on Twitter and their portfolios. And I think it was something I kind of eyeballed that I was like, okay, this is the kind of aesthetic that we very clean, very modern.

26:31 That's cool. That's awesome. I love it. These are great challenges that you've described and honestly, this conversation has been so fun because we're capturing a lot of what's happening right now with your business, I think is where a lot of people's heads are at right now in their own business, which is how do I make these key transitions early on so that when I'm five years, 10 years down the road, I can look back and say, Boy, that decision to go straight to developers and be middleware was the best thing we ever did. Or this UI interface we had designed out of Europe became what we still anchor our brand on today. These are really cool things you're describing to us. I appreciate you taking the time to do that with us. This is very cool.

27:15 Yeah, for sure. I mean, in my old company, one of the things we did, we did our UI and spent a lot of time on that. And then we actually got like a cover photo on a magazine of just our like dashboard. And so it can, I mean, it can end up being part of your marketing scheme. I mean, if that becomes your brand recognition.

27:32 Love it, love it. Well, you know, I've talked to a lot of people and I know that most CEOs and especially founder CEOs, they rarely get to where they're at, even being three years into business and still growing is a really awesome place to be. Has there been someone in your corner, someone like a group or a person or who's been there for you to kind of guide you through this whole transition from being CTO to CEO?

28:03 Yeah, no, I think, I mean, one thing I can kind of point back to is when I was at Microsoft, the team I was on at Microsoft Research, I mean, the whole team was stellar. And I've kept in touch with a lot of them. Some are, I mean, VPs at Microsoft, some are CTOs. And it's just, I think when it's sort of a meta answer, but it's, I mean, when you have a group of people that you sort of realize, okay, these are, this is a group of people that I need to keep in touch with over the years. And I think, that's an area where you're gonna find these places in your career that, I mean, you look back on, you're like, okay, that was a peak. Like, just for some reason, that was like a family of people that I work with, and everybody went on to do incredible things. And so I would just try and identify those and make sure not to lose those connections.

28:43 I love it. And that's great advice. And for those listening who may or may not have that group, find one. Because if you don't already have that group of people who could be there for you, I gotta imagine there's been a lot of phone calls, a lot of screen shares with these friends that you're like, dude, this is where I'm at. What do you think it still needs, right?

29:12 Yep, yep. And I think in more real time, the flip side of that is, I mean, don't sleep on Slack communities or discords. And I think there's a lot of real time, like even I just did a talk and met folks through a Slack community. And I mean, it went to a conference, we sponsored it, met a lot of great people and I would have never gotten to do that without joining this one Slack community. So I think that's kind of the more like new world kind of way of doing it. But it's definitely been developed.

29:44 I love it. Kirk, this has been so fun. And honestly, I love the conversation. I love your business. I cannot wait to see where you're at in three years, to be honest. Three years is such a fun point to kind of analyze and look, re -shift, and then move forward. And so I can't wait to catch back up with you later. We're gonna put links to all your stuff below. For those of you who are as nerdy as we've talked about and enjoyed what he's doing and how he's doing it. I highly recommend you've got demos, things like that we can look at.

30:15 Yeah, yeah, just on our links from our webpage and we have full documentation and sample apps and everything. And so it's free to try, free to play with and love to talk to anybody. Even just want to learn more.

30:28 Love it, love it. Kirk, thank you so much. And we will catch up with the rest of you in the next episode. But Kirk, your time is valuable. Thanks for being here today.

30:38 Appreciate it. Thanks for the opportunity.