[00:00:15] Ghazenfer Mansoor: Hello and welcome to Lessons from the Leap. I’m your host, Ghazenfer Mansoor. On this show, I get to sit down with entrepreneurs, founders, and business leaders to talk about bold decisions, pivotal moments, and innovative ideas that ship their journey.

This episode is brought to you by Technology Rivers. At Technology Rivers, we bring innovation through technology and AI to solve real world problems. If you’d like to learn more about us, head over to technologyrivers.com and tell us more about your project. Today’s guest is someone who has not only helped launch software in space, but also help countless SaaS products back to earth, running faster, safer, and smarter.

I’m joined by Todd Merrill, a seasoned software executive fractional CTO and CISO and true technologist with a career that spans innovation, cybersecurity, and product turnaround. I’ll let Todd introduce himself. Todd, welcome to the lessons from the leap. Tell us more about everything about you and then we’ll go right in.

[00:01:17] Todd Merrill: Yeah. Hi, Ghazenfer, it’s always great talking to you. I know it’s gonna be a good day when we’re on the phone. So you’ve been a great partner for me on a lot of different projects. So Todd Merrill , fractional CTO I’ve had a lot of experiences in the C-Suite both as a partner founder CEO raised money, started lots of companies, exited to, you know, private equity done some real estate along the way and then I’ve kind of settled on my niche as the software startups silverback. 

So if you look on LinkedIn, but love helping people start companies and you know, these days it’s a lot about ai, you know, and I realized earlier in the year, you know, we as technologists have had a lot of AI tools at our disposal, but there’s a lot of people struggling with what I’m gonna call practical AI or you know, this generative or LLM type Ai, in the rest of the C-suite and it’s starting to get accessible to most people, but it’s still a little tricky and I think a lot of people need just a little bit of help from a tech buddy. So that’s been kind of a fun thing I’ve been working on lately.

[00:02:38] Ghazenfer Mansoor: Oh, cool. Thanks for sharing that. You have stepped into many fix it CTO roles.

[00:02:46] Todd Merrill: Oh yeah.

[00:02:47] Ghazenfer Mansoor: What’s your playbook when you walk into a turnaround?

[00:02:53] Todd Merrill: Yeah, so turnarounds are tough. Usually there’s chaos, hair on fire. I call it a DD driven development, probably the worst case of it. Man, I didn’t even want to say the name of the company, but I was brought into after I had run a startup as CEO. A venture capital company brought me in to do a turnaround as CTO and everybody quit except for one engineer and then he had like three bosses, the CEO, the marketing guy and the sales guy and they all had opinions. He was basically just contact switching the whole day. 

You know, Hey, fix this, I gotta do a demo and then, you know, my favorite playbook is get in between that poor guy and in management and say, okay, you know, time out. We can’t do everything, but we can do anything. We just have to do it in order and then so that begins the, you know, even if it’s just a whiteboard, it’s a product roadmap. Prioritization and then just get a little headroom for the engineers and then, you know, start shipping on a consistent basis and let everybody know, hey, there’s a priority, you know, what your priorities are.

You might have to wait a little bit and that’s just the way it is and communicate early and often and you know, start hiring a team and then off we go. That was a good four year run, you know, we got the company sold. So that was a good outcome. But for about a year there, six months, it was pretty airy, yeah. So usually when there’s a turnaround, there’s a problem and you just never know. 

[00:04:41] Ghazenfer Mansoor: Oh, cool. So everybody’s talking about ai. AI is changing everything. So that’s one of the, I would say most talk about topics nowadays, and it’s becoming too much as well. Where do you see people are taking it wrong? They understand AI one thing, but in reality it is different.

[00:05:12] Todd Merrill: Yeah, so I see this huge gap where, you know, the C-suite and the investor class and the kind of the more theoretical strategic thinkers or see the potential of operational leverage that AI can bring, you know, so there’s a lot of stories about people going faster, more efficient, using ai and then, you know, from that not the CEO, but the rest of the C-suite down to the tactical level I think there’s a lot of confusion because AI’s supposed to be this magic bullet thing and people don’t know how to define it, and they don’t know how to run it and they don’t understand where this 30 to 300%, 10 x, you know, efficiency gain is gonna come from and I think a lot of what happens is you know, two or three years ago, like Microsoft for example, is pushing co-pilot, which is a little bit underpowered, it’s a good point solution, it helps you write your, you know, turn paper a little better, your email a little better. Quality of English is going up, you know, so that’s worth something, but that’s not, you know, gonna move the Enterprise needle.

So, you know, I think that’s one anti-pattern that I see is just go buy copilot or Chat GPT licenses for everybody and hope for the best that almost certainly is not gonna work out on a grand scale, you know, it’s gonna make people do a couple of things a little better web research, a little better, you know, formulation of papers and text, but it’s really not what we’re looking for, I think what has changed this year for me was the existence of this new protocol, MCP model, context, protocol that anthropic put forward, I dunno, like almost a year ago, Christmas timeframe which allows us to now bring SaaS tools, SaaS APIs into the conversation on these desktop LLM models that people have access to and for me, that’s been a huge game changer because you know, generative LLMs are great, they’re creative, they’ll fill in the gap. 

They’ll arrange the words on the paper that sound very eloquent. But then when, a lot of times when you read ’em, it’s just complete bs, you know, or just like you just making this stuff up, you know? but my CCPs kind of solve that problem, or at least a lot of that problem where you can ask factual questions against your TA traditional data repositories, where a lot of your business logic lives and then get authoritative actual numbers to questions that are interesting questions and then you make that look beautiful and I think that’s a huge improvement and then for me, I didn’t really get on the LLM train for a while and I actually didn’t get chatGPT until this calendar year just ’cause I didn’t understand, you know, what it was good for. I think MCP is kind of the big unlock here. Unfortunately it’s been kind of hard ’cause it’s really wonky and technical to set it up.

So it hasn’t been accessible to like, say, a CMO or a CFO in a large, you know, in a broad sense. But I think it’s getting better and better every day. Philanthropics done a pretty good job with that. You see an open Ai with their agent, you know, generator that they just came out with you could start to hook up MCB connectors.

I think that’s gonna be a huge unlock for a lot of people. But it’s such a new way of thinking about business, it’s just difficult for people to get their heads around until you get in there and kind of get your hands dirty and mess with it. You know and you get a couple projects under your belt and then I think once that light bulb goes off, people go nuts and then get really productive for a couple weeks and then kind of calm down and, you know, it’s fun to watch when that happens, but it’s a weird time, which we

[00:09:38] Ghazenfer Mansoor: no

[00:09:39] Todd Merrill: Yeah,

[00:09:40] Ghazenfer Mansoor: no, that’s a good mention of MCP so we call it , MCP is the API for ai.

[00:09:49] Todd Merrill: Yeah, absolutely.

[00:09:52] Ghazenfer Mansoor: Which is really exposing your API models your software into the APIs to your software into the ai so you can, anybody can just query, do anything  configuration is tricky, but I have seen like personally, if somebody can do make or Zapier, they can also do this one. Probably. I felt it easier than make, for sure. But for some people, maybe depending on who that person is, maybe I’m a techie person, so I’m able to do it, not everybody. 

[00:10:26] Todd Merrill: Well, and you’re seeing people, you’re seeing SaaS vendors start to get it and then start to bake that MCP into their SaaS UI. Right. So like, we did one the other day for one of my clients. It was active campaign. You know, they run email campaigns and then active campaign, you just log into a website with your credential and then you go to settings, you know, like you usually would get like an API token, and then it, one of the options was, you know, gimme an MCP link for my data, you know, and then so that was actually pretty cool.

And then you can plug that in as a remote, MCP, you know, just cut and paste to get that going and then it was a little weird and quirky, but wasn’t too bad, just like rough around the edges. You had to log in on a web browser at weird times. But then once we got it going we connected Google analytics for the webpage and, you know, along with that active campaign and we just very quickly mashed up those two data sets. 

With natural language, asking basic questions, you know, how many people responded to the last, how are my email campaigns doing? And it would list ’em and you go, okay, on that first one, how many people clicked on the link, you know, call to action and then how many of those people bought? And then you start to kind of, you know, paint the picture really quickly in the way. You know, software people, this is no big deal ’cause we’ve had these tools forever. Marketing people have never had those tools, you know, ’cause you had to have Python or something.

Right? And then what’s an API? And I don’t get it and, you know, to heck with it, it’s expensive but now it’s cheap. If you can kind of get past this weird little setup time, it’s an exciting time. You know and it’s kind of like you know, these old corporations like I worked at BellSouth, which turned into at.

We had, you know, a product but then we had like an internal IT department, which there’s a couple of folks that were in the back room that would bang out business apps for you to make, you know, certain processes go faster. So if you’re a big enough company, you used to be able to have an IT department that would give you these apps or you know, small pieces of software that are business custom.

I think we’re kind of all the way back around to that where people individually on their own laptop. Can create that type of piece of software just talking to their computer, if you can set this stuff up correctly and so it’s, you know, everybody had their own personal IT department that could go run queries for ’em all the time.

You know, how does that change your world? And I think it’s. Very positive in it, you know, is it gonna do all of your job for you? No, I don’t think so. I think it’s another anti-pattern, you know, hire an agent, name ’em, and put ’em on the org chart. So like, I have a virtual agent, that’s my CMO, you know, I don’t think that’s the way the world’s gonna go for a while.

I think it’s gonna make, I think you’re gonna need a really good CMO and then augment them with Ai and then you get a super great CMO. Still need the CMO, you know, and they can do way more with less resources in a hurry. But I think that’s the way the world’s going, right? It’s not replacing people, it’s just making people way better. 

[00:14:01] Ghazenfer Mansoor: Oh, thanks for sharing that you are a big fan of Claude. What makes it stand out for you compared to other relevant Gemini growth? GPT? 

[00:14:16] Todd Merrill: Yeah , so for me, you know, so Claude is a coder and so am I. So we get along very well. Claude does not do well with pictures or design, you know, so if you’re looking to do those kinds of things, there’s probably different, better solutions for you out there in the market.

Claude does APIs exceptionally well and a Claude desktop. Claude code are linked, right? under the covers. So you’ll be kind of doing something interesting in the Claude Desktop, and then all of a sudden it’ll shift into code mode and it writes you a little Python script or a little node script to process, you know, your email list or whatever it is you’re manipulating.

That’s a CSV or an xl, so that’s kind of fun the first time that happens, and that gives you a little bit of confidence. It’s not just making up stuff as a generative LLM. It’s actually writing code and then trying to get the correct answer and goal seeking which is kind of fun and then you could see the work and then check the method and then I think for me, cloud code, you know, as software people, you know, every five years or so. 

There’s all these, this new, you know, no code, low code, you know, thing that comes out and we all kind of roll our eyes and go, right, sure. This is, you know, it’s the death of software. We’re never gonna need software people again and then it’s actually really good for software people ’cause they everybody creates these little. You know, apps get themselves into trouble and then, you know, we have a lot of business to clean all that stuff up. So I thought, okay, this is probably another, here we go again. 

It’s another one of those but I think this really, you know, a step up a big step forward from those low-code, no-code generators, you can do very sophisticated things in a very short period of time. If you know what you’re doing and you’re willing to kind of babysit the code generator, it’s not a one shot deal and it can’t make up stuff.

Like if you ask it to just make up, make me a LinkedIn clone, or, you know, something very big, you are not going to be happy with the output. But if you know what you’re doing and you know what to ask for and then, you know, so like me and you have managed teams, you know, large teams and detailed teams and multi-talented teams, you know, the kinds of questions we would task our teams with and the same processes that we know and love, you know, you go do requirements review and you interview the subject matter person, you know, you get the product requirements in and then you line them up for the business requirements. You could do all that with AI or aided by AI. I love getting a transcript in, of a recorded message where we’ve had a of detailed interview and then you go do planning, and then you go do architecture analysis.

Like you would, you know, talk to an architect and then you can go through architectural review and then once you’ve made some selections and only then you start biting off little pieces of the code and you get exactly what you want and I think that’s a deadly way to do it. You still have to know what you’re doing and it, but for a very talented person, it can make them super awesome.

10x a 100x, 1000x talent, you know, more efficient. But in the same way it can, if you don’t know what you’re doing it’ll 10 or a 100x. A very poor person. So whereas you can kind of like get through, you know, now you’re gonna be really dangerous and waste a lot of time. 

[00:18:14] Ghazenfer Mansoor: I can resonate that a lot. So we are seeing those kinds of projects as you said, like people are creating it now. It’s creating opportunities for the developers. So now there are different types of outputs that are coming. So it’s like somebody gives you a one page requirement broad versus somebody’s giving you a very specific one.

So depends on, so people who can elaborate, they are thinking, well, they can articulate things. They are the ones who are able to produce those well. So, and there are some who are just giving a very high level requirement, like a clone me LinkedIn. You get something, but then you just struggle.

So there is a process of creating those as we create. So, I mean, it could be you just give initial instruction that gives you a prompt to generate enough. So sometime in our process business, what we do is we use different tools for, let’s say, UX pilot ly to even generate some screens and then take it to lovable or replic and get to the next step. 

So it’s a combination of different things that you, and it’s like building a product then. So it’s no longer, it became as easier, these tools are available, but people are thinking, okay, I can just create anything. But that also opened up opportunities these people are creating. But beyond that, now, how do you take it to the next level?

But the good thing is, instead of having a requirement doc over. It’s probably in our, the way we look at it, it’s replacing your initial blueprinting or the requirement flows part because now customers can articulate and then able to produce the initial flows. Now that’s giving a clarity to build a real product.

[00:20:09] Todd Merrill: Yeah. I’ve kind of taken this to an art form. So it started Father’s Day. My dad is a president of a Florida Condo Association, and they just passed a bunch of laws that are sunshine laws where, you know, all these, there’s I can’t remember surf side or I can’t, some condo in South Florida got cheap and didn’t maintain things and, you know, crumbled and then they caused, you know, death and then. So they made laws saying, Hey, we need to be transparent with our dealings.  So I said, Hey dad, I wanna try something. I’m gonna show you what Claude Code can do and Claude code, by the way, is underneath lovable and reptile and a lot of those really popular things.

So it’s all kind of the same code, same tools at the bottom but, so there’s this kind of process I’ve settled on where I’ll interview you in a recorded line. As the subject matter or the entrepreneur or you know, the person who needs a product for an hour, half hour or something and I’ll ask you really good questions about whether you want this or that.

Is this important or is that important? Have you thought about these things? And then take that, pull it and pull a desktop and then write, do a bunch of research and say, ask me clarifying questions. I think that’s an important thing to do, you know, because generative AI is good about, you know, generating lots of possibilities.

And then, you know, I care about that one, I do not care about that one and then go through and then say, okay, at the end of all this, I want a business plan. I want research on competition. I want pricing go to market and then I want a phased implementation approach.

And then I want an architecture, proposed architecture that could get us to market as quickly and efficiently as possible and then that’s kind of been a magic thing. You know, you can get through that in a day, maybe a couple hours and then you can go back to that entrepreneur and say, here’s a piece of paper that’s the business part of this.

This should be very close to what we talked about. Are you willing to go to the market with this? You know, this phase one and could you sell 10 of these to your buddies? If I gave you this product and then usually the answer is yes or maybe negotiate, but we can then take the output of the technical and the business, put it in alaw code, and then code it up, you know, and deploy it within a couple days, maybe a week, you know, and then get people in a business, you know, it’s a simple one, but it’s clear thought and it’s truly MVP. I mean, man, what do you, what a great time to be alive where, you know, we can get people going in a hurry for not very much money or time, you know, and of course there’s gonna be a lot of development behind that.

But, you know, if you’re in business and you could prove there’s customers, you know, much more likely to get funding, you’re much more likely to be able to bootstrap, for example you know, you’re much more likely to go borrow money too, you know, fund the business or get friends to put in money, you know, it all gets a little bit easier rather than, you know, working 20 months and, you know, I heard a guy said, Hey, I worked for 20 months. I’ve got most of an MVP done. We just need a hundred thousand dollars to finish it and it kinda, you know, two years ago that wasn’t so crazy. You know, this year that’s like absolutely absurd. Why don’t we flush it, start over and be done in a couple weeks?So it’s just mind bending how, how efficient these new coding tools have made everything.

[00:24:05] Ghazenfer Mansoor: Oh, that’s true. Absolutely. So you just started talking about this investment. So, and you are an investor as well. So as you’re talking to these different startups, what are the expectations from the investor side and what kind of startups are, I would say, easier to start or harder to start nowadays. What do you see in a startup and investment world?

[00:24:32] Todd Merrill: Yeah, it’s harder, it’s easier and harder all at the same time, you know, I think there’s a lot of different market verticals that can benefit from, you know, that can get started and get investment, right now everything’s so tight, when we talk about the whole value chain of private equity and venture and individual angels, but there’s really no forgiveness in the market right now. 

There’s no mulligans, there’s no do-overs, you get one shot at it and then you either get investors or you don’t and then because it’s so easy to start companies and generate an MVP. That’s just table stakes now, and a lot of companies do get started. 

So it’s so much tougher now because there’s so much noise at that initial seed stage, you know, getting that first check in the door that’s not, you know, your, the founder’s pocketbook, you know, and then people get frustrated because, hey, I did a product, I’m in business, you know where the angel’s at that’s not the way it works. Right and then, you know, I went to the Southeast investor round table down at Georgia Tech, like, I don’t know, a month or two ago and the Risu For which is a very active angel, you know, first check Angel organization.

The guy stood up in front of us and said, we don’t. He said he has had so many investments that didn’t exist that, you know, like 8, 9, 10, 12, 15 years old. He doesn’t wanna look at anything unless you’re gonna exit in three years and you only need one check to do it and then ideally, you know, you gotta get to a million dollar run rate within a year and I think about like how crazy hard that is and that’s the first check and then he wants to be the only check-in. 

So like, that is an incredibly hard thing to do, even for experienced entrepreneurs who know what they’re doing. You know, so many things have to go exactly right and if you’ve never done it before and you don’t have a track record and you don’t have a lot of personal capital, it’s very difficult to get anybody’s attention. It’s almost insurmountable. So that’s kind of what I’m seeing. In the early stage, you venture will back you if you have a pretty, it’s almost like a banking transaction. You know, you definitely have revenue, you know, they wanna see that million dollar run rate is kind of a magic number for a lot of investors where you’ve solved a lot of things, product market fit.

Can you sell it? Do you know how to tell people what your product is? Here’s our product. Have you built it? There’s less venture money going in because there’s less private equity money or, you know, other limited partner money going into these venture funds, you know, the interest rate, you know the interest rates on the cost of money to borrow has gotten more expensive because we have higher interest rates relative to, you know, five or 10 years ago with the zero interest rate policies you know, so it’s just, you know, it’s hard for the venture people to raise money to put it into your company, you know, so a lot of these funds are drying up, a lot of the big funds that have been successful are attracting all the money and then, you know, they’re only gonna put money into things that are sure things.

So it’s this kind of crazy hard time right now and then the private equity people, I think, you know, saw opportunity, the private equity, there’s a lot of money dry powder, but nobody wants to deploy it unless they know that they can exit, you know, in a reasonably fast timeframe. So you, there’s not been any IPOs for a while.

So I don’t know how this stuff unrolls, but you know, I think, it all kind of comes back to you need to start a business that throws off cash so you’re not dependent on the VC cycle, you know, for  money right now  which ironically is gonna make you super attractive to the VCs and the private equity people, you know, so you know, the best time to get a loan is when you don’t need it.

[00:29:20] Ghazenfer Mansoor: Absolutely.

[00:29:21] Todd Merrill: Yeah. It’s a tough world out there. So yes, it’s easier. But it’s also harder because of the money situation.

[00:29:30] Ghazenfer Mansoor: Yeah. I mean, obviously at the same time, because you can build things quickly, so the expectations are much higher as well.

[00:29:36] Todd Merrill: Yeah. 

[00:29:38] Ghazenfer Mansoor: Okay. So are you involved in any startup as well, or obviously as part of your fractional work, you work with a lot of startups. Anything interesting that you can openly share with this for?

[00:29:52] Todd Merrill: Yeah, I mean that consult is one of ’em. I’m involved with K-O-N-S-U-L-D, so they’re doing, I guess it’s like you know, AI research assistant for medical clinicians, doing some really innovative things with indexing all the whole world of PubMed and scholarly articles and YouTube, a lot of interesting medical things and then helping doctors to. You know, with their cases to try to be the medical assistant that doesn’t ever sleep and then creating community around doctors. 

So, you know, for me it’s that perfect blend of yes, it’s Ai, but it’s helping really smart people to be even better, you know, so it’s not replacing doctors, it’s making the already existing smart doctors even better and then providing them a way to talk to each other and have that hand in the loop element of all this, which I think is super important. 

So that’s a good example of practical ai, you know, I’ve got a couple others that are a little bit earlier stage probably not ready to talk about those. But you know, there’s all kinds of really innovative stuff going on, you know, in cybersecurity healthcare compliance. I think that the MSP business model, you know, in the IT world is getting ready to get disrupted big time, you know, it’s legal stuff going on anything to do with words. I think there’s a lot of noise in the marketing world. Everybody and their brothers building a couple of lightweight apps to follow up after your sales call, help you prospect.

You know, run your CRM, generate ever more clever LinkedIn outreach and, you know, cold emails that whole space is interesting because people starting to get num, people have gotten num to AI generated everything, you know, you just look at your mailbox just clogged with a bunch of stuff and it’s hard to kind of cut through that so I don’t envy the marketing world right now. So yeah, there’s a lot of really cool stuff going on.

[00:32:24] Ghazenfer Mansoor: thank you. Any latest fun projects? Your personal projects? I know we were talking earlier. 

[00:32:33] Todd Merrill: So yeah, this is I don’t know if I mentioned that process, I call it the velocity process.So I don’t know what to do with it, but I’m trying to figure out, you know, how to use it for good in the world. But you know, so we were down at the beach last week and you know, Beth does crochet and we’re sitting around. I said, Hey, let me, she’s struggling because she wants to make these blankets for people with like a college logo or a picture of  their kid or dog or something graphical designs and she has to count like these scan lines that go across and I said, Hey, Beth, you know, why don’t we just write you a little piece of software to do that so you don’t have to struggle with the, you know, how many blacks and the reds and grays and she said, okay. You know, so I interviewed her, put her through the process, and then like the next more you know, the business process basically said, if you’re never gonna make it more than $200, literally $200 a year if you’re a top 10 seller in this category. So there’s no chance in hell you’re gonna make money at this but we said, okay, what the heck? You know, it’ll make Beth happy. 

So, wrote the software and like, I don’t know, two or three hours on Saturday morning. Got it. Pretty good, you know, and then it matched kind of what she was already working on, so it’s kind of fun, you know, is that ever gonna make any money? No. But you know, is it gonna make a couple of nice people happy with crochet patterns? Yeah. Probably. You know, so is that worth doing? Yeah, You know, I guess , but that’s kind of fun.

[00:34:15] Ghazenfer Mansoor: Yeah. You always need those projects to obviously to create that, keep that fun.

[00:34:20] Todd Merrill: Yeah, for sure.

[00:34:22] Ghazenfer Mansoor: So, while we are on a personal can you share more about the podcast that takes from the skyline? So tell us who’s it for?

[00:34:32] Todd Merrill: Yeah. It’s a podcast about business consulting and venture investing and we have guys like you who are getting out there in the world and making it happen and get your stories. So you know, we at, I’m a partner at Tech CXO. In the product and technology practice area and so we realized that you know, part of being a little bit older is you have a lot of really cool friends that have done some amazing things and you know, we don’t often get a chance to tell our stories or people don’t ask us for our stories, you know, it’s always kind of intense interaction when you’re on the job, you know, some of these turnarounds and stuff. 

So what a cool opportunity to just get our friends and our business partners and you know people within our orbit, in the venture community to tell the stories about what’s going on in the world and we talk about ai, you know, just about every episode ’cause everybody’s trying to figure it out. Right and you know, we talk about the venture cycle and the economy a lot .We talk about a lot of cool little businesses that struggle for seven or 8, 9, 10 years, and then they’re an overnight success.

And that’s all everybody remembers. It’s a double digit, triple digit million exit. You know, it was easy, right? And nobody hears the story about the struggles and the trials and tribulations and you know, how it almost blew up. But they kept going anyway, you know? So those are great stories to have, you know, so it’s kind of fun. yeah, so I appreciate everybody that comes on and, you know, we try to get those stories out there in the world, you know, so I dunno how many people listen to that but for me it’s super satisfying to, you know, sit down and talk to people.

[00:36:30] Ghazenfer Mansoor: Yeah, I see those clips on the social media, so I do listen to those clips so, and obviously those 30 seconds, one minute, all those clips are really helpful. You don’t have to spend a whole lot of time.

[00:36:45] Todd Merrill: Yeah. So, well, you know, it’s fun. My I have a producer, James who’s our, he was a project manager for us on the technical side, and he’s just happened to be in LA and he just happened to be an old video producer and he raised his hand and said, Hey, I would love to do that.

So, but James sat there in the background and I just, we let him pick what’s interesting and it’s always interesting what he picks, you know, as an interesting subject ’cause it’s not always kind of what we find interesting or what the speaker, you know, would find interesting. It’s always kind of fun, you know, hearing from a third disinterested third party. You know what’s kind of interesting, for those questions.

[00:37:27] Ghazenfer Mansoor: Yeah. I was at one of the podcasts and the guy was doing something similar to what Jeff, but all, he was doing it. He was also really like spitting those messages right. During the podcast on LinkedIn? because it was a LinkedIn Live. That was interesting. Like, like picking up and then sharing those along the way. Probably expected 20 different messages out of that one 

[00:37:54] Todd Merrill: While he’s doing a Live.

[00:37:56] Ghazenfer Mansoor: What, well, I mean, he had a person, like you had James.

[00:37:59] Todd Merrill: Yeah, but how cool is that we live in a world, like it used to be really difficult just to have a video signal, you know? And then YouTube kind of democratized that a little bit and then, and now we can do all these wonderful things. You know, you can get like realtime transcriptions and five languages and you know, clips and, cartoon images and memes and you know, post it, social media, like just. Let it go. What a, I mean, software’s gotten really powerful here pretty, you know, last couple years.

[00:38:34] Ghazenfer Mansoor: Yep And Todd, so you’re a pilot as well and you’re a track lead. You have been a scout leader. So anything you can share, but then I think more importantly, what do these pursuits teach you about how to lead in the business?

[00:38:51] Todd Merrill: Oh, wow. Any lessons. Yeah. So, you know, triathlon, I did my forties and then kind of COVID kind of got me, and then I gained about 50 pounds, and then that really got me.

So I don’t do a whole lot of triathlon, but you know, it’s the whole idea through triathlon, you know, translates into business, startups are a marathon. They’re not a sprint, you know, and it’s not a lot of what you have to do in a startup or in business in general is grind work, you know, it’s the daily training, getting up outta bed when you don’t feel like it, you know but also celebrating the wins along the way, you know, however small maybe made a little personal record, you know, in your five mile run. 

Or maybe you placed on the podium in your age group, you know, or some small victory like that it’s important to mark the wins and then triathlon’s a very, you know, personal sport. So I think you know, scouts is kind of the antidote to that, you know, scouts is all about leadership and teaching the next generation, as adult basic skills for sure.

But leadership, it’s all about leadership development and then you know, kind of watching young people make mistakes, you know, in a controlled environment and to figure it out and then kind of knowing when to step in course. Correct and knowing when to kinda like nature, you know, teach physics is a great teacher, you know, so if they’re gonna burn their dinner or they forgot an ingredient in their dinner, you know, maybe that’s a good teaching moment.

You know, as adults maybe we can backstop that with a little peanut butter in the back pocket. But, you know, helping people develop their leadership skills is an eye-opening experience, you know, and it makes you personally a better leader and a better teammate too, I think, you know, how knowing how to work in teams, knowing how to plan large scale, you know, events, month in, month out those are all very valuable skills, you know, in piloting, it’s all about the getting out there in the world and making stuff happen. So, you know, Taylor from the Sky Lounge was the podcast. I think going, I’ve been able to fly to see clients a couple different times, and every time I’ve done it, it’s been paid off in spades, in unexpected, meaningful ways.

You know, sometimes it’s giving somebody a ride, sometimes it’s being able to show up, you know, last minute when you couldn’t otherwise do it for lunch or dinner or the key, you know, business meeting but it’s always worth it to get out in the world and just show up and I think, you know, in the Zoom culture that we have now, you know, we don’t even know what an office looks like anymore. You know, it’s super valuable to have those skills to be able to show up and talk to people, you know, one-on-one or in a group, or, you know, in-person setting. I think that’s very valuable.

[00:42:04] Ghazenfer Mansoor: Maybe we should do our next podcast in person.

[00:42:06] Todd Merrill: Heck yeah, let’s do it. You’re in

[00:42:08] Ghazenfer Mansoor: maybe in the plane.

[00:42:10] Todd Merrill: Okay. Yeah. I’ll  yeah, we can do that. You’re in DC and I’m in Atlanta. Both those places have wild airports, so maybe we’ll pick somewhere in the middle.

[00:42:21] Ghazenfer Mansoor: And we are right in front of the airport,

[00:42:23] Todd Merrill: so, okay. All right. Yeah, so we’ll make that happen, maybe I’ll fly up to Potomac, that’s on the south side. We can fly in there.

[00:42:34] Ghazenfer Mansoor: Awesome. Looking forward to that couple more before we wrap up. So, what opportunities do you see for the AI?

[00:42:43] Todd Merrill: It’s such a wild time right now. I think everybody’s struggling hard to figure it out. I think everybody’s got a taste for it and knows that it’s powerful, but it also does some really knuckleheaded things a lot of times and so what do you do with that? Right and I think that’s the opportunity and then so what I’ve, I’m telling people is crawl, walk, run and crawl is personal productivity and so you have to get claw desktop chatGPT, you know, Gemini, maybe grock, I don’t know about grock, but something right and start playing with it. Then, you know, learn something new and stretch and, you know, even if you don’t get it, . Today, you know, you’ll get it pretty quick and the tools are always getting easier.

So, but you do have to get on that learning curve with the rest of everybody right now and just hold on and just, you know, chip away at your personal productivity. You know, make something that will follow up on an email after a phone call. Make something that will research a podcast for you. Make something, you know, chip away at it and then make your life a little easier and then I think Walk is for me is you know, we have, once you get a couple people inside a company that get it and get what it can do you know, get a little technical help and then we can help you with team projects, so maybe go rent software, a SaaS software that has a little AI in it.

And they bolt that onto some kind of internal process, one or more internal processes that you have and they play with that and then, you know, run is enterprise readiness. At some point we’re gonna have a million, billion agents run around that everybody’s gonna create. Some are small, some are more complex, some are, you know, have a lot of working pieces.

But there’s gonna need to be enterprise security. There’s gonna need to be financial discipline. We’re gonna need uptime, you know, we’re gonna need reliability. All the things that traditional IT people have had to engineer in for forever, you know, a lot of that’s lacking right now and it’s really easy to kind of brush past that.

But if you’re really gonna run your company on some of these things, you’re really gonna need enterprise discipline and that you’re gonna need professional help, you know, for that and it’s the same people that have always done software. You’re not gonna get rid of the software people, their job.

Our jobs are gonna become very, a lot more complicated. We’re gonna need to use AI in our own way to keep up with all this stuff to make stuff, you know, better, faster, stronger to test, to comply. You know, do all the software scans to find all the bugs and fix ’em and, you know, wrangle all the policies and procedures just like we always have is set faster.

So I think crawl, walk, run, you know, personal team enterprise is a good framework and then don’t worry about being behind the curve we all are and then it’s just a wild time to be alive and AI is not going away. You know, our Roombas are not gonna, you know, take over our households and eat us, or turn into the Terminator anytime soon.

But they’re still gonna, you know, sweep our floor and make our lives a little better and I think that’s, you know, augment really good hands with AI is a good way to think about it for now and we’re a long way away from replacing hands with any kind of AI.

[00:46:31] Ghazenfer Mansoor: Yeah and I think that’s one of the mistakes.Also though people make, they assert that, oh, AI will do everything. So the hand part is definitely the most important one to consider and then a good point about the crawl part. I think what has helped with tools like Cloud Desktop, now we call it like a pre MVP phase in many of the projects were customers, any of the AI projects where customers are expecting something. 

So, okay, can we do a small mini project where just using your data, how do we generate those prompts so that his output and if you are okay with that output, then putting a tool around that is not as complicated, rather than building something and then keep experimenting and looking at it.

So we also push customers towards building that kind of a pre MVP. A tiny MVP to really see the output that you want. So, because at the end, that’s the final output that will be expected. So what is that? If we can figure that part out separately, that would save a lot of time and money, obviously. 

[00:47:45] Todd Merrill: Yeah. Experiment small.

[00:47:47] Ghazenfer Mansoor: Yeah. What advice would you give to a startup founder who feels stuck between growth and chaos? 

[00:47:54] Todd Merrill: It’s, yeah, I know the traditional one, but you’re doing it right. Yeah. I wanna hear your perspective. Yeah, it’s just the tension, you know, it startups are messy and chaotic in it. So if I like you and you’re a good client and you’re in that spot, you’ll get the accidental Empire speech so there’s this old book. They studied Microsoft, Oracle Sun, Cisco, you know, all the old Silicon Valley companies that you know, or high growth in the internet era and then, so what they found was there’s kind of three phases that companies go through kinda like an army. So you send in the commandos, it’s a small group of people with a very focused goal, one goal, blow up the tower on the hill, I don’t care how right. Arm ’em to the teeth, spend money. 

That’s kinda like the startup world, right? You just get an MVP out, you know, start selling it to people any way you can. It’s all good and then at some point you’re successful and then you have to shift gears in the second phase, which is the army of invasion. So you know, around 50 people you start need an HR person.

There’s probably some lawyers hanging around. You know, you need health benefits. There’s probably a SOC two audit lurking around somewhere. So you have cybersecurity. You have insurance because you got that big contract and you know you need that, there’s coordination and infrastructure problems.

You can’t just wing it, right? You have to have plans and then the last part is you won the war. Take the bullets outta the guns you’re at and just don’t screw it up. We don’t wanna do anything new, right? Like keep everybody on a dial up, you know, forever. Well, so the three phases you have to understand, there’s three distinct phases of operation for a project, and then therefore a company that’s made up of lots of projects.

And you have to put the right people, PE people tend to like one of, and only one of those phases and you have to understand that the people in the first phase, the commandos, are gonna hate life when they get, you know, to the steady state of the business. They’re just not gonna want to be there and then the people in the middle are builders. They can’t do the chaos. It’s their job to take the chaos and smooth it out into policies and procedures and to build infrastructure, you know, corporate infrastructure and then scale it up, right? So I think that’s, you know, if you’re experiencing scale and chaos, you’re kind of shifting gears.

You know, just acknowledge that you might need a different skillset. You probably will need a different skillset, which might mean different people. The people that started the company might not be the right people anymore, or you might need to find a better job for them, you know, in this while you’re growing and putting, put ’em on different other forward looking things, start a different other thread, a new project.

I think that’s a good way to think about it, you know, if you get to a company that’s like IBM, that’s super steady state and they’re just optimizing spreadsheets for profit. You know, that’s a much different, you know, people worked at IBM for 35 years. Like that’s unfathomable to a startup, you know, commando, where it’s like, you know, 18 months and they get bored, they want to go do something else. You know, it’s just a different mentality.

[00:51:30] Ghazenfer Mansoor:  Yep. Cool. So if you had one magic and you could instantly fix one thing about how companies use AI today. What would it be? 

[00:51:46] Todd Merrill: I think it’s change management is the hard part. There’s such a variety of experiences and expectations with what AI is, how it works, what it should do, what it shouldn’t do. There’s a lot of fear, you know, AI’s gonna, is horrible AI, you know, uses too much electricity. The business model isn’t gonna work for sure. 

They’re all gonna fail any day. Now you hear all the fear, uncertainty, and doubt. You know, there’s weird expectations about what people think it should do if we could wave one and get everybody level set on expectations for what it is, how to implement it, and then kind of what the standard playbook is, which we will do in the next couple three years. I think that would be a super helpful thing, you know, and then it’s, that’s kind of my thesis is that’s what the world needs right now, more than, you know, just throw an AI at the wall and let’s see what happens.

We need to, as technology leaders, help people through this period of uncertainty. and then figure out what normal should be and then help people change to that operational norm. You know, what does the world look like when we all know what an agent does and can agree on that?

And we have agentic infrastructure within our companies and, you know, everybody’s happy about it financially, security, infrastructure, uptime, deployability,you know, access, all that stuff. Gotta little work to do.

[00:53:25] Ghazenfer Mansoor: So we have been talking with Todd, Merrill. Todd, where can people find you if they want to learn more about you?

[00:53:34] Todd Merrill: Yeah, talesfromtheskylounge.com , is my podcast. It’s got all my socials on their LinkedIn. I’m on there just about every day. You are happy to, you know, say hello and my inbox is open. So, send me a message and if I can help you in any way possible, I will.

[00:53:58] Ghazenfer Mansoor: Thanks, Todd. This was packed with gold from architecture and AI to leadership and resilience. I really appreciate you sharing your stories and hard one lessons with us and for those listening, if you want to hear more about Todd’s insight, check out his podcast that Todd just mentioned.

Tales From the Sky Launch where he dives into business consulting and venture capital. Also, you can look up him on his LinkedIn. We will share those links also in the podcast description as always, Thanks for tuning into lessons from the leap. If this episode sparked any ideas or gave you clarity, don’t forget to share it with someone in your network who needs to hear it. 

Until next time, keep leaping, keep building. Thank you.