Oshri Cohen is an expert technologist with over 20 years of experience in growing and founding startups, and challenging the conventional wisdom of early CTO hires.
Many founders are advised to invest heavily in a CTO from the outset, often before securing their first paying customer. Oshri offers a fresh perspective on this approach, emphasizing the risks associated with premature hires.
Oshri is a CTO who knows more software and coding languages, frameworks, architectures and technologies than he cares to count. He is an ex-executive at multiple startups and ex-founder of his own custom development shop.
He is disrupting the popular thought that startups need a full-time CTO and could save on average $400K per year in direct technical leadership costs through the use of a Fractional-CTO, which is his primary business.
Oshri is also disrupting the technical recruiting business through the use of recruiters that are also engineers and can evaluate candidates on a deep technical level before recommending candidates to clients and only deal with fixed-cost mandates.
Apart from that, Oshri is also involved in what he calls a “Dev agency representation service” where he acts as a “lawyer” of sorts for companies who need to hire outsourced development agencies.
After 23 years in technology, Oshri has developed the ability to spot industry-spanning business problems a mile away.
Oshri has developed a reputation as a problem solver, leading many startups and Fortune 100 founders and CEOs to keep him around as their technical right-hand.
Oshri takes pride in providing impartial and unbiased advice and guidance, more than just your traditional consultant.
What you will learn
- How a $100,000 investment in a startup led to a tough lesson in business
- The challenges of transitioning from a successful consulting business to navigating the startup world
- How the 2008 financial crisis forced a return to employment
- Understand the concept of a “fractional CTO” and how it can transform businesses
- Discover the evolving role of a CTO across different stages of a business
- Gain insights into the minimalist approach to tools and systems
- Learn why AI is seen as a force multiplier in tech, not a replacement for human intelligence and creativity
- Plus loads more!
Transcript
Jeff Bullas
00:00:03 – 00:00:41
Hi, everyone and welcome to the Jeff Bullas show today I have with me, Oshri Cohen. Now, Oshri is a fractional CTO from Montreal Canada. That doesn’t mean he’s good at fractions, but I’m sure he is. We’ll learn more about what it is. But Oshri is an expert technologist in more languages, frameworks, architectures and technologies than he cares to count. He’s originally born in Israel, but from Moroccan descent. And that’s why he doesn’t look very Canadian to me, but that’s fine. And that makes him think he’s actually a criminal trying to escape from international intrigue. But anyway, that’s fine.
Jeff Bullas
00:00:41 – 00:01:25
So ignore that. That’s got nothing to do with Oshri at all. But we just, it is what we talked about before we got on here. So he is an executive at multiple start ups and ex founder of his own custom development shop. He is disrupting the popular thought that start-ups need a full-time CTO and we know how much they cost and could save on the average $400,000 per year in direct technical leadership costs through the use of a fractional CTO, which is his primary business Austra is also disrupting a technical recruiting business through use of recruiters that are also engineers and can evaluate candidates on a deep technical level. And as we know, a lot of entrepreneurs are shit at technical stuff, they’ve got the vision. But so
Jeff Bullas
00:01:26 – 00:02:04
apart from that, Oshri is also involved in what he calls a DEV agency representation service where he acts as a lawyer of sorts. So companies need to hire outsource development and he’ll let us know about some juicy horror stories. I’m sure about that. He loves fractional CTO World allows me to work in healthcare one moment and logistics another. After 23 years in technology, Oshri has developed the ability to spot industry spanning business problems from a long way away. So he’s developed a reputation as a problem solver leading many startups and Fortune 100 founders and CEOs to keep him around as their technical right hand. Welcome to the show, mate. It’s an absolute pleasure to have you here.
Oshri Cohen
00:02:05 – 00:02:18
Thank you very much. I am. I feel very welcome. By the way, can I hire you to explain to my mother what I do? Because I just heard my own story and I’m like, wow, that’s kind of cool. She doesn’t understand what I do.
Jeff Bullas
00:02:18 – 00:02:29
No. Well, my mother didn’t, doesn’t, well, my mum and dad didn’t understand what I did. I tried to explain it slowly but then I did it in reverse and they still didn’t get it. So, anyway, we, they, I guess
Oshri Cohen
00:02:29 – 00:02:31
As long as we’re not living in their basement.
Jeff Bullas
00:02:31 – 00:03:15
Yeah, that’s right. Well, yeah, they may think I’m just traveling the world, just messing around really. Which is essentially what I would, you know, quite like doing. So it’s good. So, let’s go right back to the future. Where, where did this all start? Where does this interest in tech, where you, you know, essentially from what you’ve told me is that you started, you taught yourself to program at the age of 13 and still did all the things that a 13 year o year old does. And, uh, we know what teenagers are. They’re just plain, really naughty and get into mischief. So, uh, yeah. Yeah, exactly. So, uh, we won’t go there just yet. We might go there later just for a bit of fun.
Jeff Bullas
00:03:15 – 00:03:20
But she made me do this one day with a gin and tonic and a beer. That would be actually more fun as well. So when I visit,
Oshri Cohen
00:03:20 – 00:03:25
when I visit Australia with my family, how about that? I’ll, I’ll, I’ll send you an email.
Jeff Bullas
00:03:25 – 00:03:33
So, what got you intrigued and doing coding at 13?
Oshri Cohen
00:03:34 – 00:04:05
Oh, my. Ok. So this is an interesting one. Vigilantism. Is that a good answer or should I? And I, and I stopped right there and I don’t give you any more suspense until like episode four, you know, season two. vigilantism. So, so back then when I was 13, you know, the dark Web, there was only the dark Web. Really? What Web Google was barely a thing. Right. I don’t think they were even making any money. They’re barely a start up. We use the Al Al al al vista. I was gonna say Al Jazeera alta vista vista.
Jeff Bullas
00:04:06 – 00:04:07
Remember that
Oshri Cohen
00:04:08 – 00:04:58
old school school? Right. So, you know what a 13 year old kid from a poor family does? And he wants video games and movies? He downloads them illegally? Obviously. Obviously that’s, that’s because, and back then we didn’t see that as I was like, hey, somebody left a copy of it over there. I guess I’m gonna pick it up, you know, surprise, surprise that 13 year old had loose morals back then. I have much better morals now. But back then, who wasn’t that interesting? Right. You’ve got, you had your napster and you had everything, right? So, um, but as it turns out people that were, were, were making that content available were masqueraded as a, real, real content like a video game or a movie that you wanted to watch with the actual title and whatnot. But they would put child pornography
Oshri Cohen
00:04:59 – 00:05:24
and that was weird. That was messed up. Um, with my friends, I went, I went to, to the police about it. Right. And they didn’t understand what the hell I was talking about. Like, when you go on the internet, what’s a computer? I’m like, what do we do with this? Right. So I learned to code because I wanted to write a virus. I knew about viruses. I was losing all over the internet all the time, you know, to the wee hours of the, of the night
Oshri Cohen
00:05:24 – 00:05:49
and it just kept on absorbing and absorbing all this information. I learned to write a virus that would destroy their machines and I would do that. And that’s how, actually I learned how to code. I started with C or C++, I don’t remember what, what I started with, but that’s how I started because I learned how to write a virus to destroy physical property. I guess that’s, I mean, Statue of Limitations has gone on that one. So it’s fine. It’s all good.
Jeff Bullas
00:05:51 – 00:05:58
So that obviously that wasn’t, it was just a, it was a hobby, but back then and it
Oshri Cohen
00:05:58 – 00:06:35
was a hobby lasted, uh, you know, we attacked maybe about 100 100 and 50 people that did that on the, on the, on the dark web. And, uh, and then we, you know, we gave it up. I mean, and it’s not our job. It was just fun, you know, for us we thought that was fun and that’s it. You know, and from there I kept on learning how to code how to, how to build websites. You know, it was not as, as, as amazing as it is. Now, you did not have the technology Right. And, uh, and, uh, right after high school I went to, I started to work, that’s it self taught. I started to work, I started to work at a small video content
Oshri Cohen
00:06:35 – 00:07:09
start up. Right. For, for a few years then I worked at a consulting firm. I had a knack for business. I had finished a business degree at the local university at mcgill University at night. And, um, and after working for the consulting firm, which uh which taught me the business of technology, which was interesting, I went off and started my own, my custom development shop and then go back to work and then my own another business, get back to work and another business and so on and so forth and just that trajectory, it was, it was, it was definitely not linear. Yeah.
Jeff Bullas
00:07:09 – 00:07:21
So tell us you, you, um, so something that was a terrible event at Montreal School, Canadian School, tell us a bit about that story.
Oshri Cohen
00:07:22 – 00:07:54
And, uh, so, yeah, that was, that was the genesis of my, of my first start up, my first real product start up. Um Basically there was a shooter at a, at a local university, a local college, ac uh college. It’s like a, it’s pre university if you for, for, for folks who don’t have colleges, um, it’s called a CJ. Uh the person was unhinged, I guess whatever the reasoning is is irrelevant. But, uh, he shot up a school. I don’t remember how many fatalities occurred.
Oshri Cohen
00:07:54 – 00:08:21
But, uh, but I have family in that school at that time, thankfully they were unscathed. Um, and we tried to call, we tried to call the school, we tried to continuously and we couldn’t, the phone system was, were little shut down because there was too much traffic and, and you gotta realize this doesn’t happen in Canada. Right. We, I think we have two school shootings ever, ever in the history of Canada at this point. I might be wrong but
Oshri Cohen
00:08:21 – 00:08:47
it’s, it, this is not a yearly or monthly occurrence. This is, if this happens once a decade, it’s a travesty. Right? I mean, it’s a travesty regardless, but still it’s not, that’s something just like that, you know, it’s not endemic in our society at least. So I came up with a solution after researching why I wasn’t able to call anyone because the phone systems were down because there’s too much traffic. But we could still text message
Oshri Cohen
00:08:47 – 00:09:13
and messaging would, would trickle in and the systems would try to continually uh try to, to, to send the message itself. So even if you sent it and they couldn’t, now they would get it eventually. Right. Um And so I built a start up where a school could notify all the entire parent body of an emergency. And then of course they had other use cases because how many emergencies do we have? We don’t have that many, right?
Oshri Cohen
00:18:10 – 00:18:54
So, I had indebted myself but close to $100,000 worth of equipment even before I had a client because, well, you know, I was a little green on the business side. I didn’t really know what was going on yet. Uh Being my first start up ever. I was self funded entirely from revenue from my consulting business, right. That I built custom software for people. And uh and I was lucky enough to get a few contracts. Right. Except the problem was the contracts were not signed yet. It’s a bit of a problem. It’s a tiny, it’s a tiny little problem, tiny little problem. Um And, and the reason why it’s a tiny, little problem
Oshri Cohen
00:18:55 – 00:19:41
um is because the Canadian or the Quebec government by the, the, my province, um, had decided to pull funding from every public education institution and even though parents would be, would be footing the bill for my product. Right. Because I presented them a business model and a, and a payment model of how they would pay for it. Right. And it will cost only a dollar a month per parent. That’s it, like, sorry, a dollar per semester per parent. There was nothing, it was peanuts. Right. They wouldn’t even feel it. They wouldn’t even know what’s going on. Um, because of that. Um, and because of the optics of buying things at a point of austerity where funding is being pulled and so on,
Oshri Cohen
00:19:41 – 00:19:51
all my contacts got pulled before they were signed. They were really basically ready to go because I had the solution ready within months of this event and it was still fresh in people’s minds.
Jeff Bullas
00:19:52 – 00:20:15
Well, yeah. So it’s, um, yeah, dealing with the government is a challenge. I have sold products and I sold, uh, personal computers to the government in the 19 eighties when the PC revolution, uh, took off. So, uh, I’ve, I’ve dealt with the government a lot and, uh, it can be challenging. Um, and sometimes it feels like a daycare for adults. Really. Uh, but that’s ok.
Oshri Cohen
00:20:15 – 00:20:22
I’m all right. All right. That’s actually, that’s actually, yeah, that’s exactly what it is. It’s a daycare for, it’s exactly what it is
Jeff Bullas
00:20:24 – 00:20:36
anyway. We’ve managed to piss off most of the governments around the world right there. And right then so, all right. So what did you do next? He went back, went back to work as an employee. Is that step back?
Oshri Cohen
00:20:37 – 00:21:13
No, I ran that. Uh no, that one failed. But thankfully my other business was going strong. My custom software development business was going strong, had many clients and so on and so forth and I had a couple of other failed products, products that I built, that failed to launch. One is, it can be closely related to Air table.com, right? What they built is what I built. I literally built that. But back in 2005 when nobody understood what that was, nobody even understood why they would, they would use a spreadsheet and database look alike
Oshri Cohen
00:21:14 – 00:22:01
on the internet. Why they, why would they ever do that? There’s no reason for it. Right. Well, there is clearly a reason because Air Table exists and they got funding and I have no idea if they’re profitable, but who knows? Right. There’s a business there. So, so I had built that, uh tried to push it. Nothing worked. And so shut that down. And now we’re creeping up to 2008, right? We’re keeping up to 2008. The financial crash happens. I have to go back to work and have to is the operative word here. Um You know, you’re independent for long enough, you start to enjoy that lifestyle, you enjoy the money you enjoy, the lifestyle, the freedom, the decision making power, you know. Um, and, and so, you know, you get used to it.
Oshri Cohen
00:22:01 – 00:22:24
Um, but I came back to work as a lead architect at one company and then the director of software development, another VP CTO VP director again and back and forth and so on and so forth. And my last job was three years ago. My last full time employment was three years ago. That’s when I went fractional.
Jeff Bullas
00:22:24 – 00:22:56
Right? So what was the look? Uh The word fractional has been around a while. Fractional CFO S fractional CTO, not more CFO S isn’t it than fractional CTO s and also the reality is that today with what we’re using right now, the technology, you know, zoom um you know, instant communication. We’ve got slack, we’ve got Trello, we’ve got collaborative software, we’ve got, you know, there’s a whole bunch of things and in your world too, I’m sure being at the right code is a lot easier than it was, you know, 25 years ago.
Oshri Cohen
00:22:57 – 00:23:21
So very, very different from when it was even 10 years ago and then 15 years ago, it changes every, every five years, there’s an entire revolution happening, but apparently we’re not supposed to, we don’t, we don’t need to learn how to code anymore. If you’re going to live, listen to the CEO of NVIDIA, you don’t need to learn how to code anymore. A I is gonna do it for you? OK, buddy. Let’s, let’s calm down, let’s calm down a
Jeff Bullas
00:23:21 – 00:23:34
little. That’s right. Step back a bit, dude. Yep. So, so where the idea for doing fractional? Um Where did that come from? Was it a, you’re working for a client and they only want you part time and you could, so how did that happen?
Oshri Cohen
00:23:36 – 00:24:25
So, throughout all of the roles that I held at start ups and even at Fortune Five Hundreds, right? Of all of those roles that I’ve held, I realized that first of all, a, they were almost all the same right in what you did. And very often the higher up you go, almost the less you have to do. It’s very, very interesting, right? How much strategy do you really need to do every single day? No, it’s half an hour on a Monday. Uh the first Monday of the quarter, here’s our strategy for the next quarter. That was more of a tactic really barely, right. Strategy is more, much, much longer term than that. Um And, and, and that’s what I realized.
Oshri Cohen
00:24:26 – 00:24:57
And then I looked at CTO job descriptions because I, I’m like, OK, well, you know, maybe I’ll look for another CTO role if this full fractional thing doesn’t work out. Right? And that just reinforced the fact that nobody knows what the heck A CTO does. Some people see a CTO as a developer, other CE A CTO as a manager, as a director as a very few. See the CTO as a visionary is a business visionary that’s supposed to bring in business ideas as well, right? And help grow the business that way.
Oshri Cohen
00:24:57 – 00:25:33
And so I saw that in myself and I saw that I wouldn’t be happy in any of these roles because I would have to be the developer there. And then I would have to be uh the leader as well, right at the same time, because they’re expecting, you know, the surgeon to be the chef to be the janitor, to be the, the nurse, the assistant and the, the guy who and the security guard, they, they, they kind of jumbled up the CTO into the CTO to every job possible, right? And it’s not supposed to be that after, after few, after further analysis and, and thoughts,
Oshri Cohen
00:25:35 – 00:26:05
um I realized that the CTO role evolves over four generations of a business. OK. Now, in your small business, a generation could be, you’ve reached from 1 million to 10 million. That’s a, that’s a nice little generation, right? Then from 10 to 100 100 to a billion, right? 34 generations, right? Depending on the growth trajectory of the of the business, your traditional business could take 50 years to get there 100 years to get there. But a start up is very different.
Oshri Cohen
00:26:06 – 00:26:53
A start up that generation can happen within six months that CTO you hired to do the development for you is now now has to be managing six months later. How do you know, you hired the right person because you hired based on one requirement, you have a completely different requirement. So you went from creative to management, which is a very difficult transition to make because it doesn’t make, it doesn’t, not ev not everyone’s happy doing that. Mhm And then from management to leadership because now you’ve got managers to meet uh to manage, right? Because your team has grown from, you know, three developers in a corner to a team of 10 to a team of 30 because now you’re getting funding right? In a very, very short period of time. Now, Jeff,
Oshri Cohen
00:26:54 – 00:27:01
what do you think? How many people do you think can make that transition that quickly and be happy to do? So?
Jeff Bullas
00:27:02 – 00:27:03
Let’s be very few.
Oshri Cohen
00:27:05 – 00:27:37
That’s the problem. It’s very, very few. And so, um and so I came up with this concept, I didn’t come up with a fractional CTO concept is very European thing of fractional CTO concept. It wasn’t really in North America, right? We like to own stuff. We don’t like to rent and borrow, right? Um And so I came up with this program, this product as a service that is my fractional CTO business, right? I would come into the company,
Oshri Cohen
00:27:38 – 00:28:21
I would reset the foundation, correct. Almost every, every problem that there is, then I would have evolved the company, I would stay with them to evolve their processes, their technology for where they want to be. Oh, you want to be what? A billion dollar company? Ok. But first, let’s become a million dollar company. First, let’s evolve you to that point. And then we’ll evolve you to $100 million.01. Right? And then help them grow into that, into that uh situation. And so I would come in, I would lead engineering. I would lead technology. I would work with every department, right? I would put in what needs to be put in the systems, the processes, I would train the people all that I could get fired.
Oshri Cohen
00:28:21 – 00:28:46
That’s my favorite thing when I tell a client, you know, perspective, clients, like how do you work? Well, I’m with you 8 to 12 months. If you know, with the goal of being fired, either I fire myself because I’m no longer needed or you fire me because I didn’t do a good job. One of those two things are gonna happen. I’d rather be the latter, right? That I’m not needed. That’s the key. That’s the key of a fractional they come in, they fix and they dig GTF out.
Jeff Bullas
00:28:48 – 00:29:01
So the, the challenge you suppose is that a CTO is a little bit like a black box. The uh the founder doesn’t really know what you’re doing except what he sees at the end, which is a functioning piece of software, for example, is that correct?
Oshri Cohen
00:29:02 – 00:29:44
Something like that? But if the CEO is doing that, that’s a really bad idea. Right. You know, under 100 employees, the founder, the CEO, the president has to be knee deep in every operation until there’s vps in there to, to handle that are, that are subject matter experts in that stage of the company. So, if they don’t know anything about tech, that’s a really bad idea. That’s a dangerous precedent. Right. And so, and so, no, you need to be there at all times. Unless you trust the CTO, you know, to the ends of the earth, it’s your business at the end of the day or unless they’re your co-founder at which point. Yeah, you can let them do their thing while you grow the business. It’s a team effort, it’s a team
Jeff Bullas
00:29:44 – 00:30:00
effort. So that’s quite often does happen isn’t like uh co founders of an organization rather tech, you got a tech co-founder and then you’ve got a visionary co-founder more so someone’s really in the weeds and the other one’s actually more visionary. Um because you don’t want them both being visionaries and you don’t want them both being tech.
Oshri Cohen
00:30:02 – 00:30:46
Uh Yes. Now both can be tech as long as as they take different concentrations, right? But you got to pick your technical co-founder really, really carefully if you pick a technical co founder because you need someone to code, right for free, make sure they can grow into the role that it needs to, that that they need to grow into because that’s akin to getting married because you want your house cleaned, right? It’s like that’s what I need you. I choose you. You, you good with broom. Come here. That’s not a really good reason because how much, how often will you need that to go on? You need someone that can grow with you. It’s a completely different ball game, especially with technology, especially
Oshri Cohen
00:30:46 – 00:30:56
because more often than not technology needs you to be hands on. Are you a hands on CTO? I’ve been asked many times. I’m like, yeah, I’m hands on, but I shouldn’t be.
Jeff Bullas
00:30:56 – 00:31:15
So do you. So what are the, if someone’s looking to hire a fractional CTO, what should the approach be? Because you’ve seen horror stories, you’ve seen great stories, maybe tell us about the process of uh hiring and look and you know, finding a good fractional CTO,
Oshri Cohen
00:31:16 – 00:31:50
you find one that has business acumen, that’s the thing. OK. Ego is completely left out of the conversation entirely. And if they talk to you tech first, you’ll have, you have the wrong person right from the get go. It should not be talking. Tech. Tech is irrelevant to a CTO. I know it’s kind of funny because it’s Chief technology officer, right? But it should be irrelevant. Whatever technology there is, we can use it, we can build with it. You have to be pragmatic and you and you, and they have to know that they’re serving business.
Oshri Cohen
00:31:51 – 00:32:26
Right. That’s the thing. Now, a fractional CTO cannot do everything themselves. It’s not possible. It’s just two hours a day. But what can you do? Two hours a day without a CTO? Can’t do it eight hours a day. Right. I get this a lot because if people think that’s what’s going to happen with a fractional CTO, they’re getting the wrong person to begin with. Yeah. And so, and so what I tell them is in two hours, I can train people and tell them what they need to do because everybody is intelligent. Everybody has a skill, everybody is smart in their own way. You have to find what they’re smart about
Oshri Cohen
00:32:27 – 00:33:04
and push him in that direction. So there’s a lot of coaching and there’s a lot of direction. This is the strategy. But you tell me if it’s the right strategy team of developers that have worked on this for the last three years and me coming in after a month, do I have it? Right? I have no idea. I think I do based on my experience. But you have to tell me and then we have a con a continuing conversation that involves the team. But what happens with that is the team evolves as well. Like, oh, we didn’t think about it this way. We didn’t think about it that way. Let me research, let me take a look, let me do this, let me do that. And then eventually
Oshri Cohen
00:33:05 – 00:33:12
they, they figure it out. Right. A fractional CTO is supposed to make themselves irrelevant. That’s, that’s what you’re looking
Jeff Bullas
00:33:12 – 00:33:51
for. Yeah, exactly. Your, your job is to make yourself get fired basically because you’ve done your job. Yeah. So in the world of uh coding and technology and software, what are some of the tools that you use today that to help with? Obviously, there’s got to be collaboration, you’re gonna be working with, you know, developers, maybe that are full-time in the business or maybe they might be part-time they might be, you know, guns for hire, whatever. But um what sort of tools do you use to manage the process and systems uh for software development as and as a CTO fractional CTO,
Oshri Cohen
00:33:52 – 00:34:35
I’m a minimalist when it comes to tools, right? There’s not, there’s not, there’s not too many I’ve managed developers, you know, 100 developers with the with bare minimal tools. You don’t need all these crazy project management tools and whatnot because I simplify how software development happens at any company, right? You may have 100 developers, but it’s really the complexity of about 10 developers because I’m gonna get 10, a group of 10 working together on a particular module, whatever may be and they’re completely independent from everybody else, even though they’re building the same product, right? So we kind of simplify it um to the point where the development team manages itself to a certain extent, right? To a certain extent. Not, not entirely.
Oshri Cohen
00:34:36 – 00:35:06
So, yeah. So I’m a very minimalist minimalist in terms of issue tracking, right, the tasks that you’re supposed to work on, right, as close as possible to the developers. So some some people will use a product like Jira or click up. I’m sure you’ve heard of click up, right? Um We’ll use whatever system is closest to the code. So usually github issues or get lab issues or whatever, it may be very, very, very close to the code because I don’t need my developers to start opening 16 windows to figure out what they’re supposed to work on. Right?
Oshri Cohen
00:35:06 – 00:35:38
Um I wanna maximize the amount of time that the developers are working, but also give them enough time to be creative. So what I do is I tell my development team and my and my clients and they’re always flabbergasted at the concept, right? Of no deadlines, there are no deadlines, there’s time caps, but no deadlines. I’m not gonna tell you you need to do this in two weeks because I’m, I don’t because I have no idea how complicated the code base will be, even though I’ll be the city of the company doesn’t mean I read every single line of code, right?
Oshri Cohen
00:35:39 – 00:36:27
Doesn’t mean that, you know, if somebody asked me what’s the timeline, I don’t know what the time, what the, what the, what the due date will be or is because that’s not how we operate because if you force someone to work within a time box and say you have to finish it by then they’re going to try to finish it as quickly as possible. And in order to do that quality suffers almost every time because I’m not gauged on my, on the quality, I’m engaged on my ability to deliver on time. Yeah. So what I do instead is I tell my clients going forward no task will ever be given if it cannot be completed in 40 hours or less. That’s it. And we expect one developer to finish one task every week in 40 hours, right. That’s it. That’s simple, right? And
Oshri Cohen
00:36:28 – 00:37:05
the definition of done here is it’s in production, it’s been deployed, it’s been pushed out, right? It’s been accepted, maybe not deployed, but at least accepted back into the, into the into the core code. And so when that happens, you have a minimum capacity that, you know, you have 100 developers. I mean, you can issue 100 tasks every week. Ok. That tells me something very rarely. Will it be 100? It’ll probably be 100 and 50 100 and 80 even 100 and 90 right? Because people are, are are are working much faster than you you expect. But you give them a minimum, they need to adhere to
Oshri Cohen
00:37:05 – 00:37:28
everything else above that is gravy. They want to do that. They want to show the quality work, the innovative because they, now they have the time to do so. So even if a task text takes 10 hours, they might spend 30 hours on it to do it properly and to do it cleanly. That’s the thing. And so it’s really interesting how you can let creativity fly this way.
Jeff Bullas
00:37:28 – 00:37:46
Yeah. So tell us about some horror stories and some success stories that you have been involved with. Let’s start with some horror stories first that you’ve walked into a company and obviously they’re going to be nameless. But, and what have you discovered? And how did you solve it?
Oshri Cohen
00:37:47 – 00:38:24
I discovered that the CTO was sitting in a closet. Ok. Doing no work whatsoever. All right. Yeah, that’s a horror story. That’s an extreme horror story. But I can give you, I’ll give you another one less extreme afterwards. Uh, doing no code, no work whatsoever. Just sitting there playing games for eight hours a day, five days a week because they couldn’t fire him because he was a, because he was a co-founder. And so in his contract, he had to have the title of CTO. So technically the CTO did nothing. They prevented him from doing any work, but in order to,
Oshri Cohen
00:38:25 – 00:38:40
but they couldn’t ban him from the office because it’s his office, his property is part owner, right? Um And when they sued him, when, when they fired him and he sued them, the judge said as long as he shows up to work, you still have to pay him.
Oshri Cohen
00:38:42 – 00:39:10
So I had to step in and be the fractional CTO. Right. Yes. And it took me four months to undo the mess that he created. Right. And make it completely irrelevant that you wouldn’t even need a CTO. I stayed with that company for eight months. That’s it. That’s as far as it goes. It didn’t take me very long because I treated every developer as an intelligent being. And I told him this is what you need to do and they say you didn’t, I didn’t know or they didn’t know what, how to do it. I would sit and explain it to them and then I’d say go study,
Oshri Cohen
00:39:11 – 00:39:53
I give you two days, go figure it out. I don’t care, you’re smart, go read, don’t do any code for the, it’s two days just go figure it out, right? This whole concept of uh of, of developers learning new tricks. I know. It’s very weird. If you take a look at job descriptions, it’s as if no one can learn a new, a new skill, then you need to know specifically this and that’s it. Yep. So, um so that’s, that’s one horror story, right? That’s one horror story. Another horror story is somebody hired me to coach the CTO, right? Uh Turns out he was firefighting all day long. He was just fixing bugs, but without fixing them just, you know, correcting the mistakes that happened because of bugs
Oshri Cohen
00:39:54 – 00:40:43
and it made him feel happy. So he didn’t really correct the issues just like what he was doing, but he was collecting $350,000 a year. Right. And managing a team of five. And I’m like, but, but, but, but like, an intern couldn’t do what he’s doing. What do you mean, you know, an intern can do this? What is this? Right. So, you know, so I coached him and, and whatnot and he ended up quitting his job because he, he wasn’t happy without, with my changes. Um He wasn’t happy with being coach. He had an ego problem clearly because if you have a problem with, with being coached, you have uh an emotional intelligence issue, you can tell you that much, right? Um And uh yeah, that’s it. That’s those, those are the two major horror stories with fractional CTO that I’ve encountered.
Jeff Bullas
00:40:43 – 00:41:09
So what are some of the success stories you’ve had with, uh walking in and um helping companies achieve the revenue with a finished project? You know, whether it’s a multiple minimal viable product, for example, which is tossed around a lot, isn’t it? In other words, creating, launching a, a piece of software to see if the concept works. Um So what are some of the success stories you’ve, you’ve um been involved with in that area?
Oshri Cohen
00:41:10 – 00:41:20
So I don’t believe in mvps or the way they’re called, they’re actually called right now is M LP S minimally loved product, I think M LP or something like that.
Jeff Bullas
00:41:21 – 00:41:24
Minimal loved product or something. I think that’s what it is. Um
Oshri Cohen
00:41:25 – 00:41:47
Yeah, something like that. We come up with weird terms because you know someone, yeah, someone at a start up that, that, that hit some revenue milestone said something. There we go. Netflix came out with micro services and then everybody started chiming back and repeating it like parrots, microservice is this and that. Um
Oshri Cohen
00:41:48 – 00:42:14
So I don’t believe in those because an MVP implies you’re just going to throw it away and you’re not going to throw it away, you’re going to evolve it from an MVP. And so I came into a company and I saw how they were coding and they didn’t care because they are, they assumed the code was going to be thrown away and as soon as they get funding, they would just rebuild it. And even the founders thought this way and I explained to them, I’m like, that’s, that’s not how things work,
Oshri Cohen
00:42:14 – 00:42:55
right? You don’t just throw away the quote where you’re gonna pause feature development for, for, for how long, how long did you expect rewriting would take? Or it will take a couple of weeks. No, it’ll take you six months to rewrite everything from scratch. That’s six months that you’re not adding new features that you’re not fixing bugs that you’re not doing anything. So in reality, you would have to have a team that’s constantly maintaining the current product while a team that’s rebuilding, that’s a lot of money. You got 10 developers on each side. Now, that’s a million bucks, at least easy if that 2 million uh for what? For zero value. So what we do is instead we evolve it and we make small changes and evolve it and evolve it and evolve it. And what I did is
Oshri Cohen
00:42:56 – 00:43:18
I recreated the foundation of their application with their, with their team of their product, right? That would allow them to scale as needed. And that’s exactly where you wanna be. You wanna be at a point where I’m growing through a credit card at AWS, increase server capacity, continue growing and continue growing and so on and so forth. That’s the point where I got them and it’s,
Oshri Cohen
00:43:19 – 00:44:16
it’s really easy to do that. I mean, it’s a joke, right. The fact that people don’t do that more often, I’m not gonna give you my secret sauce because, well, you know, gotta sell. Right. I’m in business. Come on. So, so that’s the thing. That’s, that’s the important bit, right. And I did that, they realized that within what, six months it took to, to, to incrementally improve the foundation that most of the bugs that they had disappeared because the foundation was stronger, they were able to code and deliver much, much faster, right, much faster. And, uh, and everyone was generally happier because they weren’t working on something they knew they would throw away, they had pride, which if you’re working with pride, you have your mental health and culture and motivation improves everything improves. Right? Yeah, it was, it was fascinating.
Jeff Bullas
00:44:16 – 00:45:00
So there’s one last topic I wanna cover before we wrap it up is um the impact of A I in terms of, you know, the videos CEO s going, we don’t need developers anymore, which we go, we, you’re calling bullshit on that, which I would totally agree with. So what’s the impact of A I for companies and also for developers and programmers and in writing code? Um And are, are we gonna see an acceleration in creating solutions, technical solutions because of A I I, I’d be interested in your views on A I’s role in the software and technology industry, especially the one you’re involved in as A, as a fractional CTO.
Oshri Cohen
00:45:02 – 00:45:47
So A I is a force multiplier. OK? That’s a military term, right? You inject an individual into a team and you make them better, right? You, you attach whatever resource and you make them better. So A I is a force multiplier and it’s a force multiplier of human intelligence. If you’re dumb, it’s gonna make you dumber the reality. It does go, it does go both ways that stupid question, it’s going to give you a stupid answer. I mean, that’s what it is, right? If you’re smart and you know what you know, and you know, you have good and strong foundational knowledge. A I will, will triple quadruple
Oshri Cohen
00:45:47 – 00:46:16
your capacity to produce good work. I use chat GP T and A I tools every single day. Yes, I still code for myself. I have my own projects, my own products that I’m building in between clients, but that’s what I’m doing. Right. And so the thing is that with A I is we’ve seen these A I systems that will run an entire application for you. Yeah. OK. Uh huh. Go back home. It’s not that it’s not that easy, right?
Oshri Cohen
00:46:16 – 00:46:54
The proper use of A I is write me a blurb, just a small, write me a paragraph. I’m talking about this. Just get me started. Great. Excellent, right? For code, write me a function that does this and this and this and that and this. Oh, excellent. Great. That’s what I, that’s what I got. That’s all I need. Perfect right now. I’m gonna take that. I’m gonna copy it into my code. I’m going to rewrite it because it didn’t write it properly, right? To make sense for me. But it’s helped me because I already knew what I wanted. And so I did it if I didn’t know what I want and it did it for me. How would I know it was good?
Oshri Cohen
00:46:55 – 00:47:04
So you still need developers, you still need classically trained developers and you will still need classically trained developers for many years to come. Yes.
Jeff Bullas
00:47:04 – 00:47:49
Yeah. Uh Early on about you know, the 12 months ago, there was a test done between some fashion designers that creating a, a, you know, a design and there was, they just did it for a bit of fun internally in the company. And the chief designer um was, you know, competing with everyone else because it was a bunch of designers and the chief designer came up with a much better product because they actually asked better questions and better prompts. So I’d be interested in your thoughts on the importance of prompt engineering and asking better questions that have more detail because the better the question, the better the answer I’d be interested in your thoughts on the role of prompt
Oshri Cohen
00:47:50 – 00:48:33
engineering. Yeah. So prompt engineering came uh is is is going to go away as quickly as it can. That’s, that’s, that’s kind of the thing with prompt engineering, right? It’s, it’s an intermediary technical term. While A I is getting better, you have to kind of feed it a little bit more, right? Because I can tell you, hey, draw me an image like in the style of Rembrandt, it’ll do it or draw me a picture in a canvas by paint with blotches and blah, blah, blah, blah, blah and all of these things, it will do the same thing, right? Mhm So with, with large language models, they’re, they’re going to understand what humans want. Humans want
Oshri Cohen
00:48:34 – 00:49:09
much more, much more accurately as as it evolves. And now it’s evolving at a ridiculous rate every year. We’re going to see major, major improvements. I mean, cha PD came out, what, last year in 2023 came out and, and up until then, up until now, you know, better a year, a year and a, and, and some later we have way more A I businesses. We have way more A I tools out there, we have way more everything. That’s, that’s the thing, right? So, so yeah, it’s uh it’s fascinating.
Jeff Bullas
00:49:09 – 00:49:18
Yeah. And how often are you using A I to help you code? In other words, amplify your intelligence instead of amplifying your dumbness, which you have.
Oshri Cohen
00:49:19 – 00:49:48
But sometimes I feel dumb because I can’t remember, you know how to do your basic data operation in the language and believe it or not, we google this all the time. I don’t care how good you are. You don’t remember how to manipulate dates in javascript or nobody ever remembers. So how do I do this again? I said I forgot again, right? So instead you ask judge to say, hey, I need you to create me a time, a timer function, blah, blah, blah, blah, blah. And it does it for you because I know what to ask for because I know what I want
Oshri Cohen
00:49:48 – 00:50:03
and very rarely will you copy it verbatim, copy paste and that never happens. That never ever, ever happens, right? Um It’ll happen if you see code that just doesn’t make any bloody sense that with that, with no style or anything like that. But that won’t happen
Jeff Bullas
00:50:03 – 00:50:20
now, you know. So one final question I ask all my guests is, um, what brings if you got, had all the money in the world, what would you do every day? That would bring you deep joy that, you know, that you have to
Oshri Cohen
00:50:20 – 00:50:51
do? I would have a farm and I would digitize the heck out of that farm. I would have iot everywhere. I would be building machines all day long. I’d be writing code that, that monitors everything because I am lazy. But I want to get into a business that is very, very hard. So I will be the only farmer that works three hours a day. That’s, that’s the, that’s the idea, you know, clearly I know nothing about farming and I’ve probably insulted a couple of farmers right there and then who listened to this? Yeah.
Jeff Bullas
00:50:51 – 00:51:28
Well, I’ve got a very uh good mate who came out of the technology industry and his wife had an error today, um um from her father and uh quite a big one and he’s used technology everywhere down to watering systems. Um So, you know, it’s, he’s taken his context and put it into the technology context and put it into the farm and now runs a very efficient operation. So there you go. Maybe you need to. Uh, but then you just got to buy the farm first So, there you go.
Oshri Cohen
00:51:29 – 00:51:35
I don’t know what’s real estate prices in Australia? Cos in Canada. It’s ridiculous. I wouldn’t buy a farm here. We have winter. Well,
Jeff Bullas
00:51:35 – 00:52:07
we, well, what’s happened with farms in Australia is, and then we’re right around the world, especially places with a lot of land is that, um, we’re getting the industrialization of farming, which basically is adding technology and scale to farms that are. And in Australia we’ve got, uh, farms, the size of small European countries because we’ve got the land. But, um, yeah, how long is a piece of string? Basically a farm can be, you know, five acres, 100 acres, 6000 acres. We could go to,
Oshri Cohen
00:52:09 – 00:52:26
you know what, I’m not interested in that. I’m interested in growing good food but not working very hard for it. That’s it. That’s all I do. You know, how much money does one need in life? $200,000 a year, 300,000 a year. That’s it. To live a very comfortable life. A couple of vacations, like a nice house. And that,
Jeff Bullas
00:52:26 – 00:52:46
and that is a very good statement because, uh, I think we need as human, especially, um, you know. Yeah, we both live in very highly developed countries, very wealthy countries. And how much is enough. And I think that’s what we’ve got to continue to ask each other and ourselves every day, don’t we? Yep.
Jeff Bullas
00:52:48 – 00:53:07
So, er, it’s been an absolute pleasure to chat mate. And um thank you for your time and uh sharing your stories, war stories. And um look forward to seeing you for a uh beer in Sydney, Australia at some stage when you uh are doing your fractional ct work while you travel. Is that the aim?
Oshri Cohen
00:53:07 – 00:53:14
That is the aim? I’m home schooling my kids and uh that’s it. So that’s, that’s what’s going down.
Jeff Bullas
00:53:14 – 00:53:17
So how do people contact you if they want to make an?
Oshri Cohen
00:53:18 – 00:53:30
So my firm is red corner dot IO. Um I’m the one who’s going to be answering your emails. I’m not that big that I need a salesperson yet, but hopefully one day soon enough and uh that’s it.
Jeff Bullas
00:53:30 – 00:53:34
Ok. So that’s ari at Red corner.com. Is that what it is
Oshri Cohen
00:53:34 – 00:53:37
at uh dot A O red corner dot IO, red
Jeff Bullas
00:53:37 – 00:53:38
red quarter or
Oshri Cohen
00:53:38 – 00:54:00
red corner, red corner dot IO because uh because.com is uh um how can I say uh dot com is a, is a movie from MGM. Red Corner. Ok. I had to go with dot IO. I can tell you, I like, I like red corner. It’s from uh from boxing, you know, in the red corner you got,
Jeff Bullas
00:54:00 – 00:54:08
right? Ok. So I thought maybe your job was to try and get people out of a red corner because you are the fractional CTO to save them.
Oshri Cohen
00:54:08 – 00:54:25
No, you’re in the red corner because the fighter is never alone, right? That’s the thing. The fighter has an entire team. The nutrition is the, the therapist, the physical therapist, the coach, the, the jabbing coach. Right. You’ve got everyone in your corner to help you out.
Jeff Bullas
00:54:25 – 00:54:51
And that’s called the red corner as in the color red. Yeah, exactly. I didn’t know that. All right, thanks, Ari. It’s been a blast mate and, uh, enjoy what’s left of your day in, uh, Montreal. Is that where you are? Isn’t it? I think. Yes. Exactly. And uh thank you for sharing your weather. Um I will pass that on and uh, but it might be quite different tomorrow, so that’s the reality.
Oshri Cohen
00:54:51 – 00:54:53
Who knows?
Jeff Bullas
00:54:54 – 00:55:02
Alright. Alright buddy, have a great evening and uh look forward to catching up in real life. Have a good one. Bye.