What's up party people? My name is Sunny Reese. It's Sunny Reese show. We're back at it again. My next guest his name is Russell Cole. He's going to share how he built 31 ventures as a solo founder testing ideas fast killing what didn't work and building parallel human without a traditional team. Russell, how you doing bud?
I'm excellent. Really appreciate the time. Thanks for having me.
Good to have you a little quiet. I don't know if something happened to the audio, but yeah, maybe you can move your mic a little bit closer. That'd be awesome. Okay, let's keep charging what is it that you're building? Why should people care? I guess what's the problem we're solving? We'll keep it you know short and sweet now, but obviously go deeper into it later. Hey, can you hear me Russell?
Yeah, mic test check.
Yes, louder definitely working better.
Okay, great Yeah, I think it's been a really crazy journey with AI as everybody's on this train ride and I I've been an entrepreneur at heart for you know my whole life, but technically operating as one in the startup world for about 10 years. And I've always had an issue with the loop of feedback when I communicate with any of the technical leaders at the organizations I've built or basically anything that I want to turn from a vision into reality requires development and people who are technically inclined and that loop is obviously now closed to just one person because of AI and that is what allowed me to just take a lot of time to dig in AI and I figured the best way to do that would be to start as many ideas if you will, maybe not ventures, but ideas and start developing them knowing that if they had any traction or success that they could percolate into becoming a real startup venture. And over the last 5 months since around or 4 months since around in May, I've been I've built roughly 30, call it 31 ventures and I would say four or five of them are really now trucking along, which is great. I've had a very strict process in terms of how I've built or made the decision to create a venture and that's what I'll share with you today. And then obviously kill them.
Yeah, I'm going to say like I've got 100 projects, but I just I'm curious to know what you consider something like venture worthy, right? Like okay, so we'll get deeper into it. but you know, let's slow down a little bit. 8 years in agriculture technology is an unusual road into AI tooling. I guess like what did the field the fields teach you that software people miss?
Oh man. That's that that's never been asked before. So I would say people who truly use their hands for their work, right? And not call it software sitting at a computer all day have a totally different way of working with software when they finally end up needing to potentially adopt it, right? And I think in ag and agritech what it taught me was to be extremely empathetic to the people you work with. to really understand their problem. I think in ag you have a lot of people who don't trust technology. And that was always a call it a battle, right? And it's not between me and the potential partner or client, it's just ag and technology. I think farmers have always traditionally been nervous around people, you know, selling them snake oil and getting burned and quite frankly, it's a really important job they do. They feed the world. They can't mess up, you know, too badly. But there are a lot of issues to fix. So I think everybody's aware of that and it's just a matter of bringing it all together and meshing it all. But the biggest takeaway is just be empathetic, you know, and truly understand your customer. I think when a founders build in a vacuum without understanding the problem, they get burned. And I think ag forces you to understand the problem or you'll 100% get burned.
Yeah, that really resonates. exactly was Aerial Informatics for row crop growers. What did farmers actually want that the technology industry kept getting wrong?
I think farmers just didn't have a clear view of what was happening in their fields, right? I mean, satellite imagery sounded fun, but it wasn't precise enough. It didn't give enough detail in the imagery that they could collect. You know, and this I specifically focused in a smaller crop range, which was broccoli, cauliflower, lettuce. billions of dollars, you know, are transacted in these crops every year just in America alone. but, you know, corn, wheat, soy, satellite could help enough, you know, and it's a bit of a different world, but where I was, you really couldn't see a head of lettuce in a photograph from a satellite image, right? So farmers really couldn't understand how healthy their fields were. They wanted to know the population count of their field. They wanted to project what their harvest could be so they can map out their throughput through the rest of their supply chain, right? And that's still a problem. Like people are solving this every day. I ended up exiting the business and moving on to other things, but I still am involved in talking to other companies on the exact same mission. It's quite amazing and sad on some level because not enough progress in my opinion has been made even though some amazing minds and a lot of money has been put towards it. So, yeah, supply chain throughput and really dialing in your crop population expectations for farmers is what I focused on.
And then you I guess raised about a million bucks and then sold to Nighthawk. What did that outcome teach you about like when to stop being independent?
I think that the reality of exactly was that the market in ag specifically agtech venture totally tanked the year that we experienced some massive flooding with our clients and we needed to raise capital to continue to exist essentially. The capital wasn't there. The market was totally dried up and we had to make a decision to really sell under pressure. Honestly, that that's how it worked out. you know, numbers that I won't disclose, we did sell for some cash and then I became an equity founder shareholder in Nighthawk. And it was a great opportunity for me to turn my knowledge and network and actual experience into something that I can move forward with. That experience then rolled into another challenge which was one of the founding member of Nighthawk had to essentially eject his position. He left. He had to take care of family matters and unfortunately that left me holding the bag. I was still very thankful for the opportunity and the path I was on. I held the bag for a little while, but funny enough because AI wasn't in full force then, this was just before I truly didn't know how to develop or code. Looking back, if I did and had AI now or then, I would have been able to build and sustain that business. It was about building AI bots for agritech, which again is starting to flourish now. but yeah, I mitigated my risk. I took some time off, took a breather from being an entrepreneur, and started a new role working within the agritech ecosystem, helping the ecosystem for a nonprofit in Canada, which was very successful.
Super cool. And then you founded a voice AI company in 2018, I guess years before this wave. like what was impossible then that's kind of trivial now?
For the geeks out there that really want to know about intent and entities and how to build a language model, I think that's what we were building in 2018, like truly building a language model. I loved it. I wasn't a developer, it was myself and a co-founder. he and I had very different skill sets. And the biggest the funniest thing I can remember is that co-founder, who was the tech lead, built a website to help people give us what we needed, which was to identify the entities and the intention of the sentences that we were constructing. This was for the restaurant industry. we were partnered with Jack Asters at the time, which is a Sir Corp company. They invested in and we were working with them. And we put up a website that we linked and promoted on Amazon Turk, which is some way to get people to do tasks for a very little amount of money, repetitive tasks, so that we can establish a workflow pipeline to actually build a language model. And we did, and it worked. for our purposes. Maybe not as great as the models today, but it was an amazing experience to see how language models were built. It was very difficult. We were underfunded. The company ended up getting sunset. yeah, the founder ended up leaving. The co-founder ended up leaving. I ended up starting another business, and that's what got me into Ag actually.
Cool. So, let's just switch gears a little bit. You know, so 31 ventures, one person, like what was the moment you realized I was actually possible rather than reckless?
My contract ended when after I took my break, my pause from being an entrepreneur, my contract ended. And I just jumped really deeply into AI, right? I figured this is the time. I have it. Let's do it. And I think just understanding that people who do not develop have a very different expectation of how to develop because they don't know how to develop. So, they're just using AI today, right? And you, myself, I don't know, people like me, and unclassically trained developers, right? Vibe coders. We just expect everything, right? And we'll take what we get. That's my approach. Expect everything, and then, you know, shoot for the stars, land on the moon kind of thing. And I realized very quickly two things. One is that I and I've always pushed developers, human developers that I've worked with, the same way, and it's created difficult situations with humans. but AI actually is quite receptive to it. And the second thing is that I realized very quickly like this was a big moment of understanding that there's an edge to be made here for people who are non-classically trained developers because when I read an output response from an AI model that says it's going to take 6 months or 3 months, X number of days, hours, time limitations, whatever, AI has typically made those comments to me. I think over the last month, maybe some new models are coming out and they don't mention that as much. I think things are starting to balance, but the direction that AI has given people using AI is very much coming from a developer's perspective and not really what AI can truly do. So, I think it's quite misleading, but if you're a developer and classically trained, you'll agree with it and then say, "Oh, Okay, this is going to take 3 months." Right? And kind of put the brakes on that idea just because the.
What you're talking about. Yeah, I know exactly cuz like when you're asking I'd be like, "Oh, we can finish this in 6 months." I'd be like, "You got 30 minutes. Let's go." And it's like, "Oh, we finished it." It's like, "Wait, is that also the I was going to tell you one other thing." One of my favorite things is you know, when you ask the engineering team like how much longer there and they say 2 weeks? It's not having to hear that anymore. It's like.
The most amazing thing ever.
Precious.
Precious. And sorry, I'll also say it also has taught me like what engineers really have to go through to get something done. As much as it's done super fast for us now, I still read and understand like what tools are being used, what calls, how it's working together, asking intelligent questions to teach myself. And it's giving me more empathy for developers, but I also think developers are being left behind. I'll say that very yeah, bluntly.
Yeah. So, I guess like the I mean, I so many questions, but like how do you decide which venture to keep and which to kill? And like what's the signal you trust most?
Yeah, it's the signal I trust most I'll start off with is if you well, a great idea. You need a great idea. I think everybody says ideas are a dime a dozen and they are, but it really needs to be interesting and you should also look at your competitive landscape first, right? You should really understand is this being done? Very simple questions. and my process actually is that I run this all through an AI tribunal. So, I have several different AI models that I run. Some are open source, some are frontier. And it allows me to just put an idea in and it gives me an enormous Well, it actually gives me a quick answer and ranks the idea very quickly based on the competitive landscape and opportunity size and if they think I can manage it. and if so, if it passes that gate, then an enormous amount of research is done. Still only takes maybe 10 15 minutes. and I have several different models produce results, put it in one folder, and then I have another prompt that runs that same folder with all the results to unify them and then make a decision and provide me with what information and understanding my gaps and whatever may exist so that I can ensure that it's a worthwhile venture to approach further. I really only implemented that after I built maybe 15 to 20 of my ventures, right? So, at first I was just running and gunning and building. I'll stop there for a second and I can tell you more about it after.
But I guess where does a solo operator with AI genuinely beat a funded team of 10 and like where do you still lose?
Well, I think it's still happening in the lab, if you will. a solo operator beating a team of 10. I've had a team of 10 before and I'm a solo operator right now. So, I can confidently say you do not need that many people to accomplish the goal that you set. I think it depends on what the team composition is, really. So, I to be a lot more developer heavy. I think now the team composition and this kind of boils it down boils things down to the most important gate that I need to pass, which I'll share is having a design partner on the customer side is 100% a requirement for me. And that design partner is a check mark that if I don't have the venture doesn't get continued to be worked on. so when it comes to team composition I've now built more people within my on my side to interface with the customer, the client, the partner designer that I work with. And it basically just reduces I'd say the developers from five to one and it allows me to deploy more capital and time towards other people and it's all in the name of learning the business and working with the design partner. So yeah, I would say that's really what shaped my decisions.
And that's fascinating. Can you go into like can you explain what Koya OS in 60 seconds? Like what does it actually do at like say 7:00 in the morning?
Koya OS has is more or less like a work surface that I can log into and see all of my ventures at a glance. It allows me to see all of the different pages for all of the ventures that gives me, you know, anything that an investor would typically want to see what the MRR is, if there's a client design partner on boarded, basically review the PRD, review the business plan, understand all the metrics for the business, the CAC, the run rate, the burn rate, the months to revenue, absolutely everything you can imagine that an investor would want to look at. I don't do that necessarily because I'm trying to get investments for these companies, but I see the way I'm building these businesses very similar to how an investor would look at the business as well. that's what's kept me sober, and it's very helpful. I've structured it in a way for the master business plan to be the source of truth, and then the figures and metrics percolate to the top of the view for me from that source of truth. And Koi OS is really just an operating system for me to continue to work on my ventures from a contextual standpoint and a development standpoint. So, there's both models within the ecosystem.
Groom Book has real clients? Like, how did those first customers actually find you?
Well, I had a design partner number one, right? Groom Book is one venture in Koi OS, and that venture lifted off because I had that design partner. We built the app, and then I started to get into my marketing flow and learning the marketing flow with AI, which was a challenge at first, but I've built a pretty good stack and trust it and continue to you know, improve it on a weekly basis. I've started to post in groups on Facebook where most groomers are. I reached out to probably 150 on LinkedIn with introductions, but I don't think dog groomers are as much using LinkedIn as they are Facebook. it was all automated and landed a couple customers that way. I really didn't want more than I could chew. I didn't want to bite more than I could chew off. I think right now where we're at with Fuwari and we called it Fuwari from Groom Book cuz Groom Book was taken. Fuwari is a group of three different groomers that have three different setups for their grooming practice. The app is in production. we're charging 29 USD per month for the app. And it's really still the learning phase, I'd say, with the groomers. And now I have three design partners instead of one. And I still think there's some room to grow on the marketing side. It's an extremely competitive space. I built this as one of my first ventures really just because I knew the design partner and I wanted to learn AI. It was less about was this the best fit for building the business.
Cool. I love the way you think, man. I I'd love to get a look at your dashboard someday. So I guess when Koya OS lands inside another studio, like who else has to agree before, you know, they switch how they operate?
You mean if there's another venture studio or.
Yeah, I assume is that is are you building this for others or is it just for yourself?
So far it's been for myself again from the perspective of running a venture studio. Just so that it's built as professionally as possible. But there is obviously a possibility for other venture studios and portfolio managers to absorb the operating system so they can get a clearer picture of how they're running their portfolio and get all the info they need at a glance. I think there's definitely an opportunity for that. I've been more excited about the ventures I'm building so that's why I haven't pushed that. But yeah, eventually there's opportunity for that expansion. It's not something I probably have time for in the next few months.
If the solo operator with AI is the real story of this decade, like what is a company even mean in 10 years?
Oh my god, that's a loaded question. What does a company mean in 10 years? I don't even know if it's possible to project what a company means in 2 years. The landscape is rapidly changing. There are enormous concerns about AI in the world right now with respect to removing employment for most people on the planet. We've just seen recently that all of the frontier model businesses have warned to slow down their development. This is a real challenge and question you the question you're asking is really difficult to answer. I think in a My approach to it is to take advantage of the opportunity in front of us as best as I humanly possible I can right now. Whether that leads to extreme fortune or extreme knowledge, I'm happy with either or both. I'm really thankful to see a lot of great businesses growing. But yeah, I have it's an uncertain world we're living in to be honest. It's I'm just really enjoying my time. I think that's the main thing to take away is build, be passionate, enjoy your time, enable others, support others, build community. I think the future is going to be very much community oriented. Without it, we will be we'll be decimated. You know, that's my thought.
Where do people learn more about you and where do people learn more about the projects?
You can find me on LinkedIn. I commonly kind of reshifts my profile to represent what I'm currently working on. Parallel human.ai is my study that I've launched just to examine how much one co-found one co-founder I call myself a co-founder because I work with AI. How much one solopreneur can handle. And parallel human AI LinkedIn and honestly Corbell Ops is my consulting company which I've been consulting other businesses on. Corbellops.com. If you know anybody who needs a hand, you know, developing their AI protocols and absorbing into their business, I'm happy to help.
Awesome and there's so many people watching. I appreciate all of you. My most popular stream today, so thank you Russell. And thank you all for tuning in and I will bring this one to an end. Hang out for one minute, Russell.
Oh.
