Navigating Startup Success: Insights from David Hirschfeld on the Launch First Methodology

Welcome to the Maker Manager Money podcast! In this episode, host Kyle Ariel Knowles sits down with David Hirschfeld, the founder and CEO of Tekyz. With over 35 years of experience in software development, David shares his entrepreneurship journey, including creating the innovative Launch First methodology designed to help software startups achieve product-market fit quickly.

Here are three key takeaways from our conversation that I believe every entrepreneur and aspiring startup founder should consider:

1. Focus on Solving Real Problems: One of David's most powerful insights is the importance of building products that solve real problems you understand deeply. He emphasized that the best startups often emerge from personal experiences and frustrations. By automating tasks or addressing challenges in your own life, you can create solutions that resonate with others facing similar issues. This approach validates your idea and enhances your understanding of the market.
2. The Value of Rapid Revenue Generation: David's Launch First methodology stresses the significance of generating revenue quickly. He pointed out that many startups fail because they invest heavily in product development without validating their market. Instead, he advocates for a lean approach—developing a minimum viable product (MVP) and testing it in the market to gauge interest and demand. This strategy allows entrepreneurs to pivot or refine their offerings based on real feedback, reducing the risk of costly missteps.
3. Embrace AI as a Tool for Innovation: In our discussion, David highlighted how AI transforms the software development landscape. He shared exciting examples of how Tekyz Corp is leveraging AI to enhance project management, streamline processes, and improve customer interactions. As AI continues to evolve, it’s crucial for entrepreneurs to embrace these technologies, not just as a trend but as essential tools that can drive efficiency and creativity in their businesses.

Join us as David discusses:

• The story behind the name "Tekyz" and how it reflects the company's mission.
• The various ways Tekyz generates revenue through contract software development and upcoming SaaS products.
• Insights into the importance of understanding customer problems over just focusing on the product.
• Pivotal moments from David's early career that sparked his entrepreneurial drive.
• The impact of AI on software development and how Tekyz is leveraging it for project management and product design.
• Valuable advice for aspiring software engineers and entrepreneurs navigating the evolving tech landscape.

Whether you're an established entrepreneur or just starting your journey, this episode is packed with inspiration and practical insights.

LINKS

• Follow David on LinkedIn
• David's company, Tekyz
• David's new SaaS product, ManageReferrals.com
• Learn more about Launch 1st 
• The Mom Test: How to talk to customers & learn if your business is a good idea when everyone is lying to you by Rob Fitzpatrick
• Follow Kyle on LinkedIn
• Subscribe to Kyle's Substack

Chapters

• 0:00 Introduction to the Podcast and Guest
• 0:10 Meet David Hirschfeld
• 1:06 Origin of Tekyz Corp Name
• 1:36 Business Model: How Tekyz Makes Money
• 2:08 Upcoming SaaS Products
• 2:59 ManageReferrals.com Overview
• 3:32 Beta Release and Future Plans
• 4:04 Makers vs. Managers: Podcast Theme
• 4:44 Early Entrepreneurial Drive
• 5:49 First Startup Experience
• 6:52 Lessons from a Failed Startup
• 8:18 High School Entrepreneurial Ventures
• 9:00 Transition After Selling First Company
• 9:50 Job Experience in Lake Tahoe
• 10:50 Starting a Snow Removal Business
• 12:09 Lessons from Early Ventures
• 12:20 Role After Selling First Company
• 13:41 Dealer's License and Trading Network
• 14:13 UCLA and Physics Degree
• 14:44 Sales Career Beginnings
• 15:52 First Software Company Genesis
• 19:54 Surprising Lessons from First Startup
• 21:55 Management Lessons Learned
• 23:40 AI Integration in Tekyz
• 24:56 AI's Impact on Software Development
• 28:08 Personal Experiences with AI
• 29:15 Professional Applications of AI
• 32:46 AI in Project Management
• 33:32 Using Different LLMs for Projects
• 34:56 AI's Limitations in Code Production
• 37:36 Advice for Aspiring Software Engineers
• 40:23 Future of Software Development with AI
• 43:41 Tekyz as an AI-First Company
• 44:46 Launch First Methodology Overview
• 45:04 Helping Startups with Launch First
• 47:07 Pre-Launch Sales and Market Testing

Transcript

Kyle Ariel Knowles: Hello there, welcome to the Maker-Manager Money podcast, a podcast to inspire entrepreneurs to keep going and entrepreneurs to just start. My name is Kyle Ariel Knowles and today's guest is David Hirschfeld. David is the founder and CEO of Tekyz Corp and the visionary behind the Launch First methodology, a system designed to help software startups achieve product market fit quickly and generate revenue first. With over 35 years in software development and having successfully founded and sold his first startup, David brings a wealth of experience. He's worked with over 70 startups developing the Launch First method to reduce the risks, costs, and time associated with launching a software company. Get ready to be inspired by David's insights. Welcome to the show, David.

David Hirschfeld: Thanks for having me, Kyle.

Kyle Ariel Knowles: Yeah, thanks so much. Where are you dialing in from?

David Hirschfeld: from Vista, California, which is just north of San Diego. So it's San Diego County.

Kyle Ariel Knowles: Beautiful.

David Hirschfeld: It is.

Kyle Ariel Knowles: Okay. I wanted to just go right into Tekyz, the company name. How did you come up with the Tekyz?

David Hirschfeld: I was looking for a URL that was short. This is 18 years ago. I wanted techies, IES, but that was taken. So this was the next best thing. And, you know, and, you know, five characters is a good URL, right? So, yeah, no play on words, techies, and it's pretty much what we are and what we do is, you know, develop software. So. It fit.

Kyle Ariel Knowles: Yeah, it makes sense. It's great branding. I love, I love the name of the company. So tell me just right off the bat, how do you make money?

David Hirschfeld: We do contract software development. So we work for startups and scale-ups developing their software applications, or we get added to existing teams, or we take projects that are struggling and in trouble and turn them around. And that's what we do. We have some SaaS products coming out this year. So we'll be starting to generate revenue from products that we're releasing to the market as well.

Kyle Ariel Knowles: Okay, and what are those SaaS products related to then? Just software development or?

David Hirschfeld: The first one is a product for managing referral partnerships. So there are products for affiliate marketing and channel partner relationships for larger companies, but there's nothing out there that is just focused on managing referral partner relationships and the referrals that you exchange with each other. So I was looking for software for that because we have a thriving referral partner kind of ecosystem of partners that we work with, and I didn't find anything out there. So we decided we needed a project that we could benchmark some of the new AI coding capabilities with. And so we use this as that, and that's managereferrals.com, which just went alpha release just in the last week.

Kyle Ariel Knowles: Okay. That's manage referrals.com.

David Hirschfeld: Yeah. Right. So if you go there and you want to, because you have referral partner relationships that you want to track, just use a Google to log in and it'll let you set up your own account without being invited. If you try to set up a password, then it won't let you do that. If you haven't been invited, unless you just log in with Google.

Kyle Ariel Knowles: Okay. That's great. And then when do you, uh, plan, how, how long do you, do you, are you going to have it kind of in, in, uh, in just an early release stage? And when do you plan to fully launch it?

David Hirschfeld: Uh, probably a few months and it's free to use now. And anybody who gets on the beta release will, it'll be free to use for life until, until we get to it, you know, and we may add capabilities that you, we charge for, but the functionality that's there. And before SAS will all be available for free for forever.

Kyle Ariel Knowles: That's awesome. Okay, we'll make sure and include the URL. So I wanted to since the podcast is called maker manager money. It comes from kind of Paul Graham's manifesto about there's there's makers and then there's managers and A lot of the friction is because managers want to schedule up, like on the software developer schedule, meetings every 30 minutes or whatever, and they need to have large blocks of time to code or to create, to draw, to write, those kinds of things. So can you share just a pivotal moment from your early career, perhaps from your time at UCLA or in sales, that sparked your initial entrepreneurial drive and influenced you to sort of be a maker and start creating a company of your own?

David Hirschfeld: Well, I mean, I had an aunt. My father was a developer and not a software developer, but building developer. And so I had, you know, sparked entrepreneurial inklings from the time I was a teenager. and I had lots of little entrepreneurial ventures along the way, even when I was in high school and in college. And so it's always just, I just like creating things. It's as much the maker kind of concept where I come up with ideas and I want to then execute on those ideas. So there are different types of entrepreneurs. There are the types that love to start things. There are the type that just love to create workflows and execution models. Right. And the entrepreneurs that are really good at the second and well, really good at both of those things are the ones that are just really good at generating revenue and growing a business that has value. So I've had many companies throughout the years. My first software company I started in 92 as a fluke. I was working, at the time I was with Texas Instruments and one of the other guys at Texas Instruments and I, we went out and started a, like a mini ERP product for people in the snack food delivery business. And there's a long Genesis story there. And despite every effort on both our parts, we ended up growing it to 800 customers in 22 countries over the next eight years and sold it to a publicly traded firm. But then after that, I started another company about five years later in the wholesale automotive world. It was software, a wholesale trading network for automotive dealers. And that one failed because I didn't follow the same playbook as I did the first time, which was create a very simple MVP, go out immediately and start to generate revenue from customer sales. I mean, we spent two months on our first product in the first go around, putting a very simple MVP together and we put it on the market and started selling it right away. That proved to us that there was a market and we started to understand what the market needed. And we just kept following that process for eight years. The second time I didn't do that, I went the route that you hear a lot of VCs push, which is create your fully fledged MVP and a pitch deck and go raise money, whether you've got revenue or not. And that model is what I followed the second time. And it was a very flawed approach. If I just followed the same model that I did the first time, that would have become a successful business too, because I was getting traction. But I did a free beta on that. By the way, I want to qualify that with managereferrals.com. I'm doing a free beta there. The reason I developed the product was because we needed it ourselves. Otherwise, I wouldn't have developed it and made a free beta available. I would have gone out and done pre-launch sales for that product. So I think it's an important qualification because people say, well, you're not really, you know, drinking your own, the Kool-Aid that you're mixing, right? And in this case, I'm not, but for a very specific reason, we needed something to be able to benchmark AI development tools with, and I didn't want to risk my customers on that. And we needed this product for internal use. And so that's why we built it.

Kyle Ariel Knowles: That's awesome. It reminds me of a few familiar with Basecamp, but that, that was an internal project management tool that they created for themselves because they were designing and building, you know, websites, clients, and then they thought, Oh, well, let's just turn this into a product. And now it's exactly. What were some of the first examples of, you know, in high school or whatever, some entrepreneurial pursuits you had? And then the second question to kind of tie your stories together, what did you do for those five years in between selling your company and starting the next startup?

David Hirschfeld: I was offered a position for a radio station up in Lake Tahoe in 1984, very early on, right? This is right out of college for me. And so I moved up to Lake Tahoe to take the job. It was the worst winter they had had in a hundred years. A week after, I was up there for three days and there was a collapse on the west side of the main highway that went through South Lake Tahoe, which blocked the road in from Sacramento and formed a mile-long lake where the road used to be because there was a huge landslide in the canyon. And on the east side of that same highway, it was where the road followed along these cliffs above the lake, the cliff collapsed, and so there was no way in. except from North Lake for about four months. And as a result, all these businesses shut down. So my job didn't materialize. I got there, and even though I was offered the job, which I was too young to realize how important having a written contract was back then, There was no job and there was no work. So I ended up with not enough money to just skate by and tell it was summer. So I went and did snow removal with shovels at local motels because the snow was crazy and they were hiring people for five dollars an hour to sit there and shovel snow all day long and there was like thirty of us at a motel. And so I worked there for two days and I took the money that I bought and I bought a shovel and I bought a ladder. And I went to mobile homes and started doing roofs for fifty dollars a roof. You know, for these mobile home roofs. And I could do two a day by myself. And then after a couple days of this, I went back down to the motel and I hired four guys and I paid them 10 bucks an hour each and I would buy them and we would all go to these, we would do a roof in about an hour and a half and we would do a bunch of them and these roofs all needed it because these mobile homes were collapsing under the weight and it was every day you'd see in the newspaper people dying from being crushed in their motorhomes. And then went to a property management company. This is all in the space of a few weeks and they started giving us all the big homes that they were managing right and apartment buildings. And so I was generating even after paying my guys and buying them roast beef. primary lunch at the casinos, which were a dollar 99 each. I was making a thousand dollars a day from these teams within a few weeks of being out of work. So I, it's kind of the entrepreneurial thing, right? Fortunately, I couldn't keep that business going past the season, but for a few months I was doing great, you know, from nothing to that. That's an example that was just like, it's an opportunity just I focused on how to execute and how to scale it very quickly, scale it to where it was, you know, generating nice revenue. And I wasn't doing the work anymore. I was just focusing on getting the next job and then making sure my guys were fed every day. And they thought I was great because they're getting paid twice as much and they got this great lunch every day.

Kyle Ariel Knowles: That's very entrepreneurial. All right. I love that example. And then so after you sold your first company, it sounds like you didn't you didn't do another business for five years. What did you do in between those?

David Hirschfeld: I was VP of products for the company that acquired us for the next three years and then I was casting about for the next 18 months or so until I decided to, well actually what I did was I got a dealer's license because I lived in Scottsdale. I thought a lot of my friends are always looking for their next really nice car but they don't want to buy new cars so I could source Mercedes and Porsches and things like that. So I got my license. That is such an awful industry, but I did get my license, which that was a huge effort because you have to go through like FBI background investigation and all that. So I got my dealer's license, and then I realized that there was no way to source cars except through these auctions. But there's all these dealerships and all of these wholesalers that have cars, but if you don't know them, you don't know who to call, you can't find it, somebody wants a car, where do you get it? So that's what, so I thought, okay, why don't I just create a trading network for dealers? So I did. And then I created the free, you know, the, the MVP, which I way built it out. I made it free. I had hundreds of dealers and wholesalers on the site. I wasn't charging. I tried to raise money and that, and I waited way too long to realize I should have just been charging from the very beginning, but I was afraid of, turning on the spigot, you know, the charging, I was afraid to start charging and then people wouldn't pay me for it and they wouldn't use the site. And that was the stupidest thing because if they weren't using the site, that would have told me enough more, more important information than what raising money would have told me.

Kyle Ariel Knowles: Yeah. So if, if people aren't willing to pay for it, then you don't have a good product market fit yet.

David Hirschfeld: Right. Exactly. You don't have product market fit if they're not willing to pay for it. Exactly.

Kyle Ariel Knowles: Okay. And so my understanding is you went to UCLA and you got a physics degree, is that?

David Hirschfeld: Yeah, I majored in physics, but I didn't really want to pursue that. Originally, I thought I was going to be an astrophysicist because I love science and I just like the whole theoretical thing. But then I didn't want that to be my peer group, people that I'm going to spend the rest of my life with. And I always had an act for sales and entrepreneurship. So I immediately went into sales for Computer Associates and I ended up teaching myself all the software so that I could do all my own demos. And the reason I did that is because I, it was a flip of the coin. If I was going to get a good system engineer that would do a demo for a client that would help me sell. And they never had good sales skills. And I felt like I could, you know, the technology to me was like, I can learn this. So I just learned it so that I did all my own demos. And, and then when I went to Texas instruments, the same thing. And I had that, I was selling a big system engineering product and I had my manager sanctioned me to go to the school for the system engineers. So I spent six weeks in Plano, Texas learning how to develop with their product called IEF. And then I could do my own demos. And then I ended up doing consulting after I left Texas Instruments because I was trained on all the new beta products and nobody else in Scottsdale, Phoenix area was. And so I could go out and, you know, do really well as a consultant. That's when I started the mini ERP product that I was talking about earlier.

Kyle Ariel Knowles: And that's the one you eventually sold.

David Hirschfeld: Yeah. And that's the one I eventually sold.

Kyle Ariel Knowles: Oh, it started as a fluke.

David Hirschfeld: It started as a fluke as a, yeah, it started as a fluke sort of as a side project or, I mean, like my brother-in-law had just gotten out of the military. He decided to go in the vending machine business, or the vending business, right? And he got vending machines, I think 10 vending machines, and placed them in different places. And so he was doing wrap delivery, and he'd have these bags of change, and he needed a way to be able to account for it and to manage the inventory. And the only products that were on the market were these big Unix expensive products that were made for larger vending operations. And Windows 3.1 had just come out. That was the first version of Windows that was really usable. And so I talked to a guy that was still at Texas Instruments. I was now consulting. And we both agreed, let's build a product. We both wanted to create some kind of software company. So we use this as a model, create a product that does inventory management and route scheduling and accountability for the money you're collecting versus the inventory you're putting in the machine, that sort of thing. Then warehouse inventory, truck inventory, turned out to be a very complex business project. The very first version was very simple, but then we learned very quickly as people started to buy it, that there was a need in this for the market. We had product market fit because people were buying it, but we needed to improve product solution fit, which was that it was providing the value that people needed. So, and then we really learned that we have stronger product market fit was that fall. So we launched it early, late spring of 92. Then that fall, we went to the first trade show, which for the vending industry, these are huge because you've got Pepsi and Coke and M&M's and, you know, huge booths and big and the manufacturers of the machines and the manufacturers. Anyway, turns out we thought it didn't realize how big a trade show would be, but it was pretty significant. We had a tiny little booth in the back corner because it was a last minute decision on both our part. A tiny little booth. We even went to Kinko's and had them print a sign for us that came out that was on both sides that said the name of our company. And we had bought some PVC, just total garage shop kind of thing, to push it out in the middle so it would stick out of our booth so you could see it from both sides of the aisle because nobody would ever find us. Literally in the back, a little 10 by 10 booth. We were buried. People found us, the word started spreading around the show about what we did and that we were the first Windows product on the market. We were inexpensive and we didn't, and we had people lining up outside the booth down the hallway, like six or seven deep so that they could watch, you know, be the next one to see a demo of the product. And we sold dozens of copies of the software We didn't even have time to eat lunch any of the days that we were there. I mean, that was, it was really blew us both away when neither of us expected anything like that. So that's when, you know, you got product market fit. And then it's a matter of racing to get product solution fit so that you keep your customers, you know, once they buy the product. And we just kept building it. And then we get bigger customers that had multiple routes and multiple warehouses. And so you have to keep building out the software to satisfy that. the merchandising in the vending machines because some products start to slow down in sales. You've got to see that the velocity is shifting and what products we're going to increase the velocity based on historical trends. And it turns into this big complex project over time and we ended up, you know, and then we ended up selling it to a publicly traded firm out of Toronto.

Kyle Ariel Knowles: That's really cool. So did you, what was your role mostly? Were you mostly sales or were you on the operation side or were you wearing multiple hats?

David Hirschfeld: Multiple hats. I was always involved in design and product management. So you can think of me as the kind of the stakeholder on the product side. My other partner was the chief architect and developer. But we had a development team that we ended up hiring over time. And we had a sales team that we ended up hiring, building over time. It just kept kind of growing, right? At first it started out as a virtual business where everybody worked remotely. And this was before any, there was this concept of remote, but they were all local in our area, but just worked out of their homes. And we had a router, you know, press one for sales, two for technical. You know, there's technical support, three for customer service. Uh, and it would just route to different people in their homes. And then eventually we ended up getting an office and everybody worked out of the office and it just, it just evolved in the, in a way that it needed to, to, to grow into a, you know, real viable company.

Kyle Ariel Knowles: And what was the most surprising thing you learned from starting, you know, founding your first tech company and selling it?

David Hirschfeld: Well, I didn't learn it until much later after I started TechEase and watched a lot of companies fail. And that's the founders that focus on revenue first and love the problem and love talking to their customers. Not the product, but the problem. Those founders find a path to success and find a path to growth. Founders who love the product and believe in that they'll be successful because they've got a great product. Those are code words for I'm going to fail.

Kyle Ariel Knowles: I love that. I love that answer. Let's talk a little bit about management then. So you successfully built and sold your first startup and looking back on the management side, what were some management lessons that you learned from your first startup and from what you're doing today?

David Hirschfeld: Always hire, that one I did learn, only hire somebody if they're smarter than you. So if I'm hiring somebody in marketing, I want them to be smarter about marketing than I am. So that my job is to get all the impediments out of their way and facilitate their ability to accelerate, you know, accelerate what we, what our objectives are in that role. If they're in a management role, they have to hire people that are smarter than them. and then their job is to facilitate them. If you hire people that are not smart enough, then you have to carry them and other people on the team have to carry them and it sends the wrong message to everybody else. And if you make a mistake in thinking you hired somebody who was really smart, but they're not, then trying to bring them along and carry them is also sending the wrong message to the team because it makes everybody else's responsibility to fill in their gaps. The other thing you learn in that is to be ruthless about getting rid of people that you realize you made a mistake when you hired them. And doing it quickly executing on what the inevitable thing that you're going to end up doing six months from now is do it in two or three weeks just do it right away because you're not helping the person that you hired by carrying them because they don't have a future at your company and you and everybody you're hurting everybody else. And it changes, it affects the morale and affects the culture where everybody wants to be inspired and raised in terms of their level by the people they work with, not feel like they're being burdened by them.

Kyle Ariel Knowles: That's a great lesson to learn as a manager. When you start techies, are you using AI to begin with, or does that come later?

David Hirschfeld: Yeah, we didn't use AI to begin with. We started Tekyz 18 years ago. AI was there, but it was really, really early, way too early. We started playing around with AI in 2014, I think, or 15, when TensorFlow came out from Microsoft, because that made AI consumable to the average developer. Not average developer, but somebody who wanted to do some R&D. It was definitely not mainstream for a long time after that. when OpenAI came out a year or two before ChatGPT, that's when it started to really become available, but still most, almost nobody knew about it. I read an article a year before ChatGPT came out, an article in Scientific America that was written by OpenAI about itself. That's what opened my eyes to the fact that it has really come along. The fact that you could talk to AI like it was a person and it would understand you and respond to you like a person. That was the first time I had actually experienced that. And then we started to get much more involved with AI and started experimenting with it in terms of generating code and solving logical problems and things like that. And then, of course, JCPT comes out, and now AI is consumable by everybody. And at that speed that it's accelerating, we've never seen anything in technology quite like this. I haven't. Some of my favorite things about where AI is going and what it's capable of right now are things I discover almost on a weekly basis. So for example, this was about a month ago, Chat2UPT had just come out with their vision capability. You know, they had the voice where you could talk to it and it would talk back. That came out like six or seven weeks before that. And then the vision part came out where if you And still most people don't know about this, but you, if you're on chat to PT and you tap the, the, the speak button where you're talking, having a conversation with it, there's then if you've got the paid version, there's a little camera button right next to it. And you tap that. And now it's looking out the front of your phone while you're talking. And it can react and see what you see. So I was in Lowe's because I was picking something up. And then my wife said, would you pick up some fertilizer for our hibiscus? And I said, well, what kind of fertilizer? Anything. Just look for hibiscus on the fertilizer. It shouldn't be any problem. I said, OK. and typical instructions right nothing says i discussed on it i'm looking at this massive wall of fertilize bags of fertilizer like a hundred different brands uh… and i could not find i discussed in a recited stood about thirty feet back and i turned on chachi pt and i said i'm looking for fertilizer. I'm holding it up like this. I'm looking for fertilizer on these shelves that would be good for hibiscus. What can you recommend?" And it said, oh yes, I see there's tonal. And I said, which one is that? He said, well, it's in a red and white bag. And so I walked up and I said, here's, I see some red and white bags. And Chachapiti said, no, it's to the left and it's on the second from the top shelf. And I went like that. So I see it now. And I walked over and there it was. And I went and looked and Looked on the back, it says hibiscus on it and so. like in just twenty seconds i didn't have to figure out and then i did it again looking for a particular uh… cough medicine in the pharmacy where i was just what i said i'm looking for certain cough medicine tell me when you see it and then by walking on walking down the aisle it says well you'll find it in a red and green box folks dot i see it as i'm just walking down the aisle it says it's on your right on the bottom shelf and i look at is right there This is like game-changing. I mean, obviously this is just simple, saving you the frustration of having to hunt through shelf, but I will never look through 50, 60 brands on a shelf again. It's just so easy to have chat find it for me instantly. And that's like nothing compared to what it'll be able to do another year from now, right?

Kyle Ariel Knowles: Yeah, that's the worst it will be.

David Hirschfeld: Exactly. That's the worst it will be, exactly.

Kyle Ariel Knowles: Well, that's interesting because I could see, you know, those are two great use cases. I know they seem simple, but again, we're living in this futuristic state where we always have someone that knows a lot more than we do right in our back pocket. Yeah, I've never used it in that way before. Now I'm excited to turn the camera on because I use the voice mode all the time. But I've never used it in combination with vision.

David Hirschfeld: Oh, it's amazing. It's just amazing. And it's fast, too. I mean, when it's not having one of its moments where it's not running fast, but most of the time. But it's instantaneous when it's running in its normal speed, yeah.

Kyle Ariel Knowles: Well, you kind of beat me to my question, but one of the questions I had for you was on a personal level and a professional level, what are some of the ways that you've been sort of amazed or just blown away by AI? So I guess those are two great personal examples. On the professional side, what are some ways that you've just been amazed of what it can do for you professionally?

David Hirschfeld: Well, you know, now it's helping us with design. And so I'm using it to design applications, you know, our own applications and make changes to it and improve the UI and things like that. And also to build our applications, to optimize existing applications that we had built previously. One of the things that we're working on right now also is a voice interface. So there's a lot of voice related stuff in AI now. And there's where you can train it on your business and things like that. But project management is a really complex process. And managing the details of a project, especially for large projects and through many different releases of an application, gets to be daunting because you end up with lots of duplicates. feature requests that get added multiple times and all this planning and scheduling and what's been done and feedback and bugs and all that. We're putting a voice front end on that with the idea that this voice bot will be an expert on that person's project, both on the requirements and existing tasks and bugs and everything else. If we're working with a client, they can say, I just found a bug, On the screen and and if it's and if the client's not giving it enough information, then the voice bot will start asking you more information to make sure it understands exactly what the client saying. And then if it finds anything similar in the system, it'll say, I think I found a bug that's similar to that. Let me know. If this is the same bug and the client says, well, it's similar, but it's missing this and says, well, I'll update the bug. Or it's the same and it'll tell them when that bug is, what the priority is and when it's scheduled to be added to a build. And the client can then say, well, can you update that to a higher priority? The AI can come back, say, yes, I will put that in the next build, but that will delay it by a day. Do you want me to move something else to lower priority? Right? All of a sudden now that our client has a direct relationship, personal relationship with the project, with the project itself from a project management perspective and can ask it questions about when's the next build, what about this bug that I reported last week, oh it's already in a current build, I don't see it, well please refresh your cache and tell me if you still don't see it. Things like that they would normally have a conversation with the developer about. So we're building that right now with the idea that we could take that same intelligent voice bot and apply it to other business problems besides just project management. And that's one of the SaaS products that we're working on that will be out this year. But again, building it for internal use. Another one is being able to do niche analysis for startups. That's one of the hardest things to do is figure out who is the early adopter niche and nailing it. So you know what the value proposition is, what the personal impact to the stakeholder is that makes them lean forward when you talk about the problem that you're solving. These are very difficult things to nail down. We have a methodology part of Launch First to do that, but it's tedious and time-consuming, and a lot of founders fatigue before they get all the way through it. So that's one of the projects that we have on our plate. Another one is how we estimate projects, right? Because whenever we're working with a new client, we do a very sophisticated estimation process. More than a lot of companies do because it's part of our value add, you know, it's doing this incredibly detailed estimate. which is very costly and time-consuming for us to do, but we always do it. So we've been working on an AI model to basically automate doing our estimation process using AI to even improve on how we do our estimates. So these are all different products internally that we've been working on using AI to build just for ourselves, all of which are needed by, you know, our cohorts in the industry.

Kyle Ariel Knowles: And what LLMs are you using for these projects?

David Hirschfeld: I'm using Gemini, Clawed, OpenAI primarily, Lama, Extract for one project. So there's a number of them that we're using. The big names that you hear about because they have the big money behind them. And plus they're with companies that at least are, have some level of integrity that we believe they have some level of integrity.

Kyle Ariel Knowles: Are you using different ones because different models give you better results for the different applications that you're using them for?

David Hirschfeld: Yeah, different results, different cost structure. There's some automated routing tools that real-time pick the best model for a particular prompt. So you may have a workflow and in different steps of the workflow, it's dynamically picking different models for each of those steps based on cost and effectiveness. So there's a there's a whole mix of things, right? So that's agentic workflows for people that know what that is, where you've got agents that are intelligent in a certain way, that are speaking with other AI agents that are have different level of intelligence in a different way. And they're working together to solve problems, or execute workflows.

Kyle Ariel Knowles: Okay. And have you found one model that's better at producing code than than the others?

David Hirschfeld: Well, to ask me if that's a weekly question, new Gemini is really good. Claude is really good. The newest version of Chachapiti, they're all really good and they all fall short in the same way. But even the places where they fall short are starting to kind of get those gaps filled. If I want to build some big sophisticated system, and have AI build the whole thing for me by giving it a set of requirements, it won't be able to do that. It can do small, simple systems like that and get most of it right. But as soon as it gets to a level of complexity, it falls apart. And the reason is, it literally forgets what was over here when its context shifts. If it's context shifts and the thing it may have actually created this thing over here on this side, right? but its context has gone like this, it forgets that even exists. A person doesn't. They remember that's there. They may not remember why they did it or what it is, but they remember it's there. And just knowing, so that level of fuzzy kind of, oh yeah, that's right, that's part of my world that I have to consider because, and I'll know because I'm a developer that it may, I may affect it if I make a change over here. And then you turn around, you'll look at that and you evaluate it, then AI doesn't do that. AI forgets that, right? Those types of problems are being slowly being handled. Well, one of them is they keep making the context bigger. But that just means it's bigger before it hits the wall and it forgets the thing out there. It needs like almost this fuzzy level outside. Like I remember all these things out here even though I don't quite remember what they do, but I remember at a very high level that they are associated with this thing I'm working on right now over here. You know, the human brain does that. The AI brain doesn't do that. It can, it can apply massive amount of information at a level of detail. That's impossible for us to do, but it can't do that kind of fuzzy. I don't know how to describe it, but the fuzzy kind of onion, onion level understanding of things.

Kyle Ariel Knowles: Yeah, that makes sense. So, so what advice would you give to someone that's going into software development now, say they're, you know, a senior in college, and they're getting a computer science degree or software engineering degree, and they want to, say, do a startup, what kind of advice would you give to someone like that they're starting now?

David Hirschfeld: If they just wanna do a startup, but they have no idea what kind of startup they wanna do, I would tell them to focus on being productive in some way. Automate things in your own life. The best startups come out of solving your own problem, just like Basecamp. Just like what we're doing internally with techies. We're solving our own problems and those are becoming products that we're gonna market. And the reason why that's, The best way that doesn't mean it's the only way, but that's the best way of addressing of creating a company is around as a problem because you understand at a cellular level. the problem you're struggling with, why it needs automating because, because it's, uh, and there's a value proposition there for you personally. If you're in an industry full of people, just like you struggling with the same problem that you're struggling with. Now you have a way of validating product market fit. You have other users like you, you can talk to. and understand how they struggle with the problem, what they've done in the past to deal with those problems, to try to mitigate the problem, how much that problem is costing them, how much it personally impacts them. Is that similar to you? Is it different than you, right? You start to triangulate and broaden your understanding of the problem that you already understand at a certain level. This creates a roadmap and a path of building a business. So we weren't solving when I started that first company, that was a fluke. My brother-in-law came to me and asked me, you know, what can you do about, do you know anything about this? Do you think you can build something for this? And so we started to understand his business from his perspective, right? And then we realized that there was a potential business here. I said, how about we build this for you and make it work for you? So we knew we were solving his problem and that gave us the basic fundamentals for what the industry needed. We started reaching out to the rest of the industry and then we realized, okay, we're solving it for him, but not for these other people because we're missing key factors that others need in general. And so we started filling in those gaps. And then you just keep broadening and expanding, focusing on the most high value things that appeal to the broadest part of the market so that you increase product market fit and increase the value proposition.

Kyle Ariel Knowles: And so I guess if someone's going into software engineering now, my question is, you know, with AI and AI being able to produce code, is there room, like there's been room in the past for software engineers and how would one differentiate themselves to actually be a software engineer going forward?

David Hirschfeld: Okay, I look at it like this. Every time there's some significant innovation, So this is our technological bubble, right? What's possible. And the people that are on the surface of this are the people that are capitalizing on it. When some new technology innovation comes out, it makes the bubble bigger, right? People that were in here are no longer viable because their skillset is not needed anymore except with legacy systems. because what's needed now is a skill set out here. That's true when, you know, if you think of, if you go back to machine language, where it was ones and zeros, right? And then assembly, where it was the, you know, moving blocks around and memory registers and things like that. And then you went to the first level of abstracted languages, like C, and then C++, and then you went to the next level of, you know, you just keep basically shifting what your skillset needs to be, right? And this bubble's going like this, right? So now people that are writing code on a daily basis, that skillset just isn't going to be needed in the same way. You still need people that understand code, right? Because you have to be, because AI can't just do it all yet. but you don't need to be the code jockey that you needed to be a year ago or two years ago. So what does this skill set look like? We're just starting to understand that. It's a blend of having coding skills and prompting skills and also understanding how to architect systems. So there's a big architecture component skill set that's developing in terms of like, how do you You conceive of agents and then weave these agents together. And so it's like the sphere is getting bigger. There will actually be more opportunity, but it's not going to be based on your coding skills. It's going to be based on some broader skill set. But it will be a bigger market because so many more things can be created. So many more things will be created in lots of different directions that couldn't have been conceived of before AI came out. There's a bigger market now for technical people. What those technical skills are is just forming now. It's not being a Node.js coder anymore. It's not that. Although having those skills is helpful and important still, but you can learn them so much more, so much more easily and so much faster because AI helps you learn them. Right.

Kyle Ariel Knowles: Yeah.

David Hirschfeld: So you've got people writing Node.js now that don't even know how to code. Right. But they're writing it because they're asking chat to be T to write something for them. I had somebody tell me they wrote a script in Google app scripts. They didn't even know what that was, but they wanted to automate something related to email, right? How emails were coming in to get rid of all the massive emails that they, that just weren't important. And so it wrote a Google app scripts and told them how to install it. And they did it and it worked. And now that was JavaScript. Right. May have never written code before. So things like that. So it's just changing really fast.

Kyle Ariel Knowles: As far as your company Tekyz, and I've read somewhere, I think maybe on the website that you're an AI first company. Does that mean that all your employees are thinking AI and doing things, or do you have a specific department or people that are focused on the AI aspects of what you're doing?

David Hirschfeld: We drive everybody to use AI. to help facilitate their jobs and to be more creative in how they execute their jobs. Everybody, right? Nobody gets to not use AI at this point because it's too much of an accelerator in every aspect, whether it's an administrative role or whether it's a testing role or whether it's a, it doesn't matter what it is, a deployment role, a development role, right? my executive role or a director of operations role. Everybody is leveraging AI.

Kyle Ariel Knowles: Really, my last question related to techies is you have a section on your website called Launch First Methodology.

David Hirschfeld: Right. Yeah.

Kyle Ariel Knowles: And my understanding is that your company uses this methodology in all of your projects. And then are you also just offering this basically free information to anyone else that wants to use Launch First methodology or creating a community around it or how do you see it going?

David Hirschfeld: So I've just seen so many startups fail because they believe in their product, right? And they believe everybody's going to come and they spend all this money building it and then They may be, and they put out free betas, you know, with the idea of creating a business. That's the only reason they're doing it, not solving. And it's one thing if you're building something internally to solve a problem, then you put a beta out to start to get feedback from users about what else they need, right? That's different than building a product for the sole purpose of creating business. because free betas are rarely successful in that world. Primarily because you're not solving a problem you really well understand from a personal perspective at that level. And even if you do understand it pretty well, you're not building it and using it on a daily basis yourself. Because that using it on a daily basis informs you about where the value proposition is, what makes it worthwhile to using it, and what needs to change if it's not usable. So often companies will spend hundreds of thousands of dollars before they realize nobody wants to buy their product, and now they don't know where to go or what to do. Panic, pivot, panic, pivot, and they fail. And I've seen it happen 95% of the time with startups that don't find a path to revenue quickly. So LaunchFirst is a methodology that's an alternative approach where somebody, we basically go through this very exhaustive niche analysis process with a client. Now we charge for LaunchFirst, so differently than if they're doing a project. We build the prototype, design prototype like we would for them, but we build it a little bit more involved because we want to make sure it demos like a real project.

Kyle Ariel Knowles: So when you talk about helping startups that are having a hard time, is this launch first to them to help them better figure out product market fit?

David Hirschfeld: If they're already building their product, it's very difficult to get an entrepreneur to step back and consider a different path, right? So which is what they should do, but it is Really difficult. And I understand you get so much momentum going in the wrong direction. It's hard to want to, you know, you know, you want to see it through because you're not sure it's the wrong direction. Right. So I understand that. So what we do is we build that design prototype. We've done the niche analysis and we build a, and, and we go and then build a marketing stack and test the market with this, with this design prototype. And with the idea that we're offering some high value pre-launch sale model to somebody who, to a potential prospect who we should know who that is at this point. And if they won't buy it, if we're offering a really high value opportunity, like a lifetime license or something like that for, you know, paying for one first year of the subscription fee in advance, and they actually pay you in advance. And if it's a big enough problem, that has a high enough cost to it being and and uh… and they are personally impacted by this problem which are all key factors for an early adopter that they know they're gonna buy the software when it comes out and you offer them a high enough value to buy in freelance sale then they'll buy in enough numbers to prove you got product market fit before you ever write a line of code and uh… if they're not buying it in this model then what's wrong with the business model because They should recognize that this is an opportunity they don't want to miss out on. But they're not recognizing that, or their problem's not big enough, or you've just got the wrong early adopter, right? Because they're not personally impacted by the problem in a significant enough way that they're willing to act on it now. And then I have had founders that say, wait a minute, you're giving away lifetime licenses. I won't be able to get any more revenue from that client. Well, compare giving away an immeasurable fraction of a percent of your market. Let's say to raise the same amount of money from sales that you would get. If you give 20% away to an investor for, you know, seed investor, that's a, it should be a no brainer, right? When you put, when you think of it like that, it's what, yeah. Plus you have all these highly invested beta customers when the product does come out because they've paid for it. Right.

Kyle Ariel Knowles: Yeah, it sounds amazing. I just have a quick lightning round of questions for you and then we can wrap up. So what's a book that you recommend the most to people?

David Hirschfeld: My favorite book is called The Mom Test. I don't know if you've heard of that book. It's my favorite business book. because it basically teaches you how to get actual valuable information from potential prospects and customers when you're thinking about creating a product. when it's typically almost impossible to get real information from them that will be valuable when you're in terms of deciding what the value proposition is and what the personal impact is of problems related to a product. But it teaches you how to focus on the problem that you're trying to solve and the customer.

Kyle Ariel Knowles: Yeah, you're the third or fourth guest I've had to recommend that book. And my son is going into an entrepreneurial program. He's a software programmer as well, developer. And he says it's one of the best books he's read.

David Hirschfeld: It is, it's really exceptional, yeah.

Kyle Ariel Knowles: Thanks again for the recommendation. So what's your favorite candy bar?

David Hirschfeld: Oh, let me think. My favorite candy bar, Mars Bar or Milky Way. Yeah, one of those.

Kyle Ariel Knowles: Favorite music artist?

David Hirschfeld: Favorite, oh, that would be yes. Favorite cereal? I don't have a favorite cereal. When I was a kid, it was Captain Crunch, but I don't eat cereal, so I don't have a favorite cereal at this point in my life, yeah.

Kyle Ariel Knowles: Mac or PC?

David Hirschfeld: I have a PC.

Kyle Ariel Knowles: Google or Microsoft?

David Hirschfeld: Oh, Google.

Kyle Ariel Knowles: Dogs or cats?

David Hirschfeld: Dogs. I mean, I love cats too, but dogs for sure.

Kyle Ariel Knowles: Phantom or Les Mis?

David Hirschfeld: Oh. Probably Les Mis.

Kyle Ariel Knowles: OK, and the last question for you, David. What's the worst thing about being an entrepreneur, and what's the best thing?

David Hirschfeld: Oh, the worst thing is that you just don't get to ever relax, right? Because you always have, you can't, it's hard to leave work at the office. The best thing about being an entrepreneur is that you never get to relax. Because you always have something new, exciting to think about. I mean, it's why we are entrepreneurs, right? I mean, here I am after all these years and I'm coming out with several products this year, launching this, which will all become their own businesses.

Kyle Ariel Knowles: It's really amazing. And thank you again for being generous with your time, David. I wish you much success with the products that you're launching and your current company techies and, and the launch first methodology. And I look forward to seeing what happens over the next few years with these new products that you launch.

David Hirschfeld: And really, thanks for having me, Kyle. I really appreciate it.