Todd Merrill:

Hi, welcome to Tales from the Sky Lounge. It’s a podcast about business, consulting, and venture investing. We get out there in the world, we talk to people who are making it happen and we get their stories.

Today’s guest in the Sky Lounge, Matt Oess. Hey, Matt. How are you?

Matt Oess:

Hey, Todd. A longtime listener, first-time caller. Thanks for having me here.

Todd Merrill:

Hey, I remember that guy.

Matt Oess:

No. I always wanted to say that.

Todd Merrill:

Yeah. Well, hey, Matt, who are you and what are you working on?

Matt Oess:

Thanks. So, I work for TechCXO. I’ve been here for just over 12 years. My practice is mainly sales and marketing consulting. Maybe half of my time is spent there. The other half of my time is spent leading our executive coaching practice. And for the last 14 or so months, life has just been an incredibly deep AI dive.

Todd Merrill:

Very cool. And I’m sure we’ll get into that. AI is on the lips of many people out there in the world, and it’s a complex subject for sure. So, let’s talk about your coaching practice. What stage of company do you typically get involved with in your day-to-day coaching?

Matt Oess:

Yeah. It runs the gamut. We target, I would say, 50 million to a billion. And it’s usually an investment made by the C-suite in order to help them get more aligned, find better trust, get decisions to stick, etc. But we also spend quite a bit of time with TechCXO clients. So, an up-and-coming leader, a founder who’s trying to figure out how to live in isolation. And as sort of the lonely place, you’re a founder, and now you’re the CEO, and you wear 1,000 hats.

So, just helping them navigate life and create better poise and polish and build leadership skills.

Todd Merrill:

I’m involved in an investment group here in Atlanta. And we have these mentoring programs. We had a conversation with one of the fellows who’s running a program about a couple of troubled mentees. And the subject came up of, hey, can you go help this other… This one client or this one company that’s in the program very early on, has a founder with a very strong personality. That’s good and bad. Right? It cuts both ways. And then we were talking about how sometimes people can take the advice and grow. And sometimes they can’t. And then they’re just not ready or it’s the wrong time. Do you run into those situations where you get on the call and go, man, you know, I’ll help you, but you just know they’re wasting their money at this point?

Matt Oess:

It’s a good question. I get that a lot. I think the way that I might answer that is coaching is for people who want coaching. Coaching is for people who are just on that good-to-great journey. And it doesn’t always have to be remediation of some perceived flaw. Right? And so there’s a lot that has to happen early on in order to build trust with that coachee and get them to understand that this is the one place on the planet where there’s no judgment. This is the one place on the planet where you can be vulnerable. And we live by this 100% confidentiality because being a founder, being a CEO, or being a C-suite executive, or even directors and VPs, it’s a lonely place. Right? So, you have to be transparent to your board but you can’t be fully open because if you were fully open, they’d probably throw you out on your ear because they’d realize you don’t actually know what the heck you’re doing. 

Todd Merrill:

Ha ha ha

Matt Oess:

And that’s not just them. That’s humans. Right? But on the other hand, you have to be transparent to everyone who works for you. But you can’t be fully open because if you were fully open to them, then they would run away screaming because they would figure out that you don’t actually know exactly what you’re doing. And I think what happens to a lot of these leaders is they have to feign this perception that I’ve got everything under control. Right? And that makes vulnerability really, really difficult. So, the coaching experience is really the one place where you can have complete confidentiality, where you can find clarity because if you find clarity, you’ll find courage, right? And that’s what you need in order to sort of move on. And so we work a lot on this skill, I call the awareness skill, which is, you know, everyone, no one shows up to work and says, you know, I think today I’m just going to do the opposite of what my gut is telling me because I just want to screw things up. Like nobody does that.

Everyone feels like their intentions are right, and their hearts are in the right place. I’m doing this because I feel it’s the right thing to do. It’s your brain that sort of gets in the way of all that and keeps people from kind of doing what they know the right thing is to do. They just can’t control their behavior because I think so much of it is subconscious.

So, we spend a lot of time in those early days, helping that founder really and truly understand how their behavior might be impacting others and maybe counter the culture that they want. But then more importantly, we can have a conversation where they can, you know, I don’t know if it’s the right way to say this, sort of be naked or like be vulnerable, right? And then let’s poke at those things and let’s figure out what’s getting in the way. So, for the rest of the world, you don’t have to put up that false front any longer.

Todd Merrill:

Lay your cards on the table, so to speak. Yeah.

Matt Oess:

So, I would say 99 times out of 100 people, you know, dive in after the first couple of engagements. They’re like, okay, now I get this because I got to be honest. If you had backed up 15 years ago, an HR or some CEO came to me and said, Matt, good news, we hired you a coach. It probably would have been organ rejection for me. I mean, you know, it’s hard because, to some degree, we’re all these A-type personalities, right? We want to demonstrate not that we’re the smartest person in the room, but that we have everything under control. And so it took me actually going to coaching school and really trying to understand what this thing really is before I realized, oh my gosh, I’ve been going to battle with two-thirds of the possible solutions and the logical part of my brain. Unlocking the other one-third of solutions is sort of what I learned in my coaching journey. And that’s the journey that our clients are on, me and the other partners who do coaching at TechCXO.

Todd Merrill:

Yeah. You know, and on this podcast, we run into culture being very important early on, right? And then I feel like this is another unlock, right? The faster you can get through your ego being in the way, I got it all figured out. I know everything to “Hey, let’s work through the issues. Let’s solve it. There’s a team here. I’m not important. The team is important.” That seems to be an evolutionary step for a lot of young leaders, particularly first-time leaders, or maybe second or third-time leaders, too. They just hadn’t figured it out yet. It’s interesting that your coaching is in the later stages of business when everybody has a little bit more resources, when there are more stakes on the line, when there are bigger problems, you know, you have to perform at a very high level, which is probably pretty elite. It’s like an elite athlete at some point.

Matt Oess:

There’s one of the things I always say about coaching, like, I’m not your therapist. I’m not your counselor, right? Those people are doctors and professionally trained, and they have helped you with some wounds in the past. And you get rid of that wound in the past so that you can be healthy and productive in the present, right? A coach is somebody who takes somebody who’s perfectly healthy in the present and accelerates their journey to a better future that they want, right? If you’re an elite athlete, you want to go to the Olympics, you hire a coach, right? If you are a good leader or a great leader and you want to be even better, you hire a coach. 

Todd Merrill:

We did triathlon for a while. I had a nutritionist, and then I had somebody that would periodize my training schedule. Then I had somebody else in the gym who would help me lift, get strong, but not bulky. So, there’s all kinds of help to go faster, go stronger, go longer. In the same way, do you focus on… you have a sales marketing background. Do people come to you and say, Hey, Matt, I got to crack the nut on sales marketing? Or is this more of a… it’s a higher-level conversation at some point, right?

Matt Oess:

Let’s see. A company that I’ve worked with in the past on sales and marketing, we’re just catching up after that engagement is over, and they’re talking about some of the things that they’re running into. And maybe that is cultural or it’s the dynamic that’s happening in the C-suites. CEOs always say I have these great functional leaders. They’re just amazing people. But when we get in the room, we fight like cats and dogs. And they’re all good people, and I don’t want any of them to go, but we have got to get rid of this dysfunction. We’ve got to get rid of this. And so we try to help them navigate that stuff and try to help them with their culture.

Todd Merrill:

Yeah. I think it’s super important.

Matt Oess:

Yeah. But then if you back up, like we’ve worked with a ton of startups, growth phase companies. And usually what will happen is… I had one group, which was an environmental company, believe it or not. And these guys created this unbelievable business. They were just printing money. And they get to a certain point and they’re literally 20-somethings or early 30s. And they’re trying to figure out how to navigate this thing. And they haven’t had the experience of a lot of us who have spent 20 or 30 years learning emotional intelligence and politics and figuring out how to deal with that.

Now you have kids that basically have created a $15 million a year company, a $100 million a year company. And they’re like, what do we do? And so they’ll come to me and they’ll say, do we need to step aside and bring in professional leaders because we don’t really know what we’re doing? Or the thing plateaus and they don’t know how to take the next step. So, a lot of times we’ll just get engaged to help them understand that they’re much more powerful than they think. And they don’t need 20 years of experience in order to figure out what are the leadership skills we need to develop as a team. What are the communication skills we need to develop as a team? And sometimes that leads to consulting like two-day strategy sessions or things to help them paint a better vision or help them with some more strategic initiatives, etc. So, it kind of ebbs and flows, you know. Somebody who hired me for coaching might hire us for strategic planning or sales and marketing, etc., etc.

Todd Merrill:

Yeah. And so it just helps to have an experienced leader in the room that you can just say, hey, I got this problem. And then you might say, hey, A, I can help you or B, I have partners or I have friends from my history that would be perfect for this kind of thing.

Matt Oess:

A lot of times they’ll run into a specific problem and we can bring in one of our CFOs or one of our COOs in order to help them solve that specific problem. But also, one of the things that’s unique. There’s sort of, I don’t know, a statute at coaching that a great coach can coach anybody. Right? And a really gifted coach is just the art of asking the right question to help get that shift and get that pivot within that client. But what I find is we have about a dozen coaches at TechCXO. And so by day, most of them are a fractional CFO or a fractional CTO, and you might be a fractional CFO for five or 10 companies, but they just also have a propensity and a skill set to do some coaching. And they sort of do that maybe 10 or 20% of their time.

What’s really unique about our offering is we might work with the C-suite and we might have the CFO and the VP of finance connected with one of our coaches who is a CFO. And so that really helps break down the barrier because, okay, well, this person is probably worthy of coaching me. So that kind of helps drop the barriers down a little bit. But then also, you know, sometimes toward the end of the conversations, I’m going to step out of my coaching role for just a second. Have you thought about trying this, this, this, and this? Because our CFOs, for example, have just seen so much. And sometimes it’s like consulting, which clients just absolutely love.

Todd Merrill:

Yeah. Awesome. Well, take it a step further. You know, your Purdue work is kind of interesting. So not only do you counsel people, you’re involved with some diplomacy. Do you want to talk about that for a second, what you’re involved with there?

Matt Oess:

Sure. So, do you mind if I share my screen?

Todd Merrill:

Oh, yeah, absolutely. That would be great.

Matt Oess:

Wish me luck here.

Todd Merrill:

Sales guy that showed up with slides. This is awesome.

Matt Oess:

Let’s see here. Okay. So, this is the Krach Institute for Tech Diplomacy. I’ve been engaged with them for about six months on their advisory council. And I’m really quite excited about it. It’s just a really neat thing to be a part of. So, it’s a dot org. It’s a nonprofit, but it’s a bipartisan institute. And the whole idea is we want to accelerate all of the innovation that’s going on in the world that we’re creating with our allies, NATO and other people that are in our trusted network. And the whole idea is to protect that technology from falling into our adversaries.

So, today that might be China, Russia, some of the Middle Eastern countries like Iraq, etc. And we just want to be really careful that we’re not taking our best technology and it’s falling into their hands. So, this organization was behind some of the things, like they were heavily involved in the CHIPS Act. And a couple of years ago, I think the world kind of dodged a bullet. And I’ll tell you that quick story.

One of the technology areas that we work on is 5G and 6G. And many of you know about Huawei. Huawei is a Chinese company. And basically, what they do is there are no intellectual property laws in China. So, they will take, for example, an Ericsson or a Motorola base station and some of the technology in a 5G, 6G network, and they’ll reverse engineer that, reverse engineer everything down to the chassis and the chips and the software and whatever. And then they’ll take those products and they’ll put them on the market for maybe 30 to 40% less than those people who incubated the technology.

Well, they were starting to make announcements that all these countries, 90 of them, were going to start buying the 5G and 6G networks. And people got really nervous because there’s no way of knowing if there’s what we call kill code in that technology. And so this organization got involved, started to work with universities and governments and companies in order to say, wait a minute, if that kills codes in there, all of a sudden your entire network goes down for your entire country. And if lots of countries do this, we’re in a very vulnerable place. So, this organization got deployed and they got 90 countries to sign some diplomacy and some policy types of things. And none of those countries are going to buy this 5G, 6G Huawei technology.

So very, very advanced. We’re very engaged in semiconductors and microelectronics. So, all the AI chips, that’s Taiwan semiconductor, etc. Just trying to be really careful about where those things go.

I am involved because of artificial intelligence and machine learning and also my affiliation with Purdue. But now that I’m into this, I’m just finding so many places to learn and sort of get engaged. And it’s really fun to actually have an opportunity to have a global impact. I am very interested in space and I’m interested in hypersonics. And obviously, I’m spending an incredible amount of time with AI, but all these things are just great to be a part of it.

One of the unique things about this is this advisory council. It’s made up of retired brigadier generals, the chief scientists for hypersonic weapons at Boeing. So, it’s a little bit unnerving sometimes when you’re in an audience with some of these people and you’re thinking, how did I sneak in here?

Todd Merrill:

Yeah, yeah.

Matt Oess:

But on the other hand, about once every couple of weeks, whenever there’s something newsworthy that comes out, they throw us up on a Team’s call and there might be 20 or there might be 50 or there might be 80 folks on it.

For example, when DeepSea came out with the announcement that they stood up for this new LLM and it was so cheap and people were worried about the data going to China. So, Monday, the story came out. On Wednesday, there were about 80 of us on a call and we’re just talking about what this means, what the risk is, what we want to do about it. But what’s really great is you actually find out the skinny about how they built it and in the marvel of the technology. So, it’s as much about policy and protection as it is just the technology itself.

Another great example is the Munich Security Summit about a week and a half ago. And so we got on a call with probably about 50 folks and its experts. They’ll always bring in experts. So, for DeepSea, it was two people from AWS, two people from Google, and some academics. And so you bring in this incredible brain power in order to talk about these topics.

Todd Merrill:

Yeah. My very early career was in telecom. And so we’ll talk about that. It was at Ciena; you were at Nortel. So, we were kind of going at it at some point in our careers. But that was a big concern. We’d spent $50 million to build these very sophisticated ASICs and a fabulous semiconductor. And then how do you audit that when it comes back? And then Huawei was a thing back then. There were some others. And then definitely some national security considerations. I got to be careful, you know. And then you have some very interesting conversations in some very dark rooms.

Matt Oess:

Yeah

Todd Merrill:

And I had a cable modem experience. I was on one of the first cable modem projects. But the boards go out to China, they get manufactured in China. And then I can remember they came to Atlanta. The engineers said, “Hey, good news. We’ve cost-reduced your board.” We said, “Hey, that’s really cool.” And then they whip out this brand new design with a whole bunch of shortcuts, and then half the really spooky parts missing. And they obviously gone in and like, it looks probably meant nothing. And then we all kind of went, holy crap, this is not good. But it’s just different, you know?

Matt Oess:

Yeah, you reminded me of a story. So, as you know, I was at Nortel Networks, and I was in our long-haul fiber optic division. And I sort of had a market development hat. And Huawei came up then, that was 2000, 2001. So, this isn’t a recent problem. This is an old problem. We had these 10 gig OC192. I know you’re seeing a guy, so you know the language. Literally, Huawei came out with their version of it. And every single detail was exactly the same, from where ASICs are laid out on a board, to the chassis, to the plastic enclosure, and even the blue LED, right? That was the Nortel Networks little logo, that sort of was backlit. They just changed it to Huawei, but it was the same color.

Todd Merrill:

The documentation. We saw the documentation where they didn’t kind of substitute, you know, Company A for Company B. And it was just like, you tear your hair out and stuff like that.

Matt Oess:

Yeah, it was really a scary time for technology and we still have to deal with that today.

Todd Merrill:

Oh, yeah. And I don’t think it went away. I think everybody got more sophisticated and the domain changed, right? So, I love that slide with all those very critical technologies on the list. I mean, it’s a consideration for sure.

Matt Oess:

Yeah. You had some space things early in your career, didn’t you?

Todd Merrill:

Yep. Yeah. So, grad school, NASA paid for my grad school, had some work, experimental work go up on the shuttle back when that was a thing. And yeah, SATCOM. SATCOM was very well-protected, you know, and we can’t go into it too much, but you know, it was good times and high security. And there’s definitely like, you know, you’re either in that world or you’re not. And it’s important to protect those secrets. And you worry, you know? I certainly reflected back on my career without naming names or companies or environments. You look back and you go, yeah, that person was definitely up to no good, you know?

So, I had … I probably shouldn’t tell too many stories, but you know, like we had janitors come in and this a long time ago, at night. And you’re like, you two are way too smart to be janitors in it, you know? And you’re just like, you’re definitely from a country like that we should be raising our eyebrows on. And then, you know, it was very, you know, cutting-edge stuff. And we were working like 90 hours a week, you know? So, we’re always there at night when they come to pick up the trash, they pick up the trash, but you’re like, but they also kind of checked out the whiteboards and everything. And then, you know, we had a manager at a different place that went to another place that’s involved in Silicon today, you know, and you’re just like, yeah, you know, they definitely grew up in a different country that is maybe on that list and then you just wonder what’s going on. And, you know, don’t know.

Matt Oess:

I couldn’t be more excited about the Krach Institute. I just think that we can have an amazing global impact and we just need to be careful when we need to be, you know, really thoughtful about this stuff. We’re creating an incredible innovation and it’s giving us a great competitive advantage to make the world a much better place. We just want to be sure that it’s not turned against us. And so you know it was founded by Keith Krach, who was an undersecretary of state under the first Trump administration. And he was also the CEO of DocuSign when they went from almost nothing to something huge. And so he’s a really, really great guy to sort of just be around, but he wields a lot of diplomacy swag, I guess, or swagger. Right?

And then the other co-founder was the president of Purdue, Mung. So really, really need to be just really privileged to be a part of it. And hopefully, I can have an impact on the globe.

Todd Merrill:

Well, and then, you know, that’s the benefit. If I’m all the way back looking for a mentor, it takes a long time to buy, you know, to gain that credibility, to be in that room at all. And then when you’re in that room as a senior leader with your experiences, and you’re contributing to that conversation for sure, but everybody else is too. I mean, that’s a massive amount of bandwidth that’s getting exchanged on some very critical key directions internationally, you know, that’s an invaluable experience. You know, you don’t see that. I don’t know where you go shopping for that on the internet, right?

Matt Oess:

Yeah

Todd Merrill:

You can’t just but let’s talk about AI. This is fun. I’m an old school, you know, machine learning, did computer vision back before it was sexy, you know. Before NVIDIA started getting involved with all this stuff, you had a really cool demo the other day. I don’t know if you’re willing to share that with us or give us a little peek, but you as a business user have a very different perspective from me as a hardcore software guy. I’d love for you to kind of go through that. And then maybe, you know, you can help me kind of unlock a few of these, you know, everybody’s wrestling with LLMs and calling it AI, right? And then why is it important? And then is it important? And then what are the limits? You know, maybe we can talk through some of that stuff. I don’t know if you got that demo or some part of it that you’d like to share.

Matt Oess:

Yeah, there are probably a couple of things that I could show, maybe just to give the users an appreciation for some of the stuff that’s happening, some of the stuff that’s coming, but maybe just to back up. So, you know, I guess, I have this hypothesis that everyone’s career runs for about five years and then there’s some kind of a pivot, like that’s kind of happened to me time and time again.

So, you know, when I came to TechCXO from having what I call a real job and doing consulting and the things that I’m very blessed that I get to do every day, cause I love waking up and going to work. Right? And so I did sales and marketing for five years and then sort of got the itch to try something else new in addition to what I was doing. And that’s where executive coaching happened. And now that’s been up and running for eight or nine years and now AI happens.

And so this AI thing has just unlocked this part of my brain that I turned off the minute that I sort of left Yale and left my MBA. And I was like, “Oh yeah, I got to turn on my engineering card ’cause I’m going to the dark side into sales and marketing.” And now this AI thing happens, I don’t know, 14, 15 months ago. And I was like, “Oh my gosh, this thing’s a thing.” It is just, I mean, literally in the last 14 months, talking about thousands of hours on top of my work week, just nights and weekends, just playing with this stuff. And it has just consumed me and I’m loving it.

You mentioned about the LLMs. I think the LLMs are really important in this fight for supremacy or to be better or faster or cheaper or like, however, they’re going to build these things. I think all that is important for the underlying foundation of the innovation. And obviously, the chips are important in order to sort of make that stuff real. I don’t even think we’ve started to scratch the surface on the value that this stuff is going to create because the value is all in the software side of this thing. And these multi-agent types of things and custom agents. And, you know, and I think we’re just barely starting to scratch the surface of that stuff. And so I pay a little bit of attention to which LLM is in the lead on this or which LLMs are in the lead of that. And I just do that because I deploy different LLMs for different things.

So, for example, when it comes to writing and creative writing, I like Claude for that, right? Now, some of the others have caught up, but it suits my needs. I don’t try a bunch of other ones to see if this one is slightly better and meets my needs until I hear about something better then maybe I’ll test it a little bit. But for coding, I might use a different LLM for just doing some analysis. I use OpenAI and GPT, etc. But my passion is really on the application side of this thing.

Like, the value is whether it be productivity or, you know, training a model to do something, to transform the accounting department of a client. And a lot of my thinking is really around … I have probably done 50 or 60 little projects with this platform that I’ll show you here in a minute. But it was just my own productivity, my own curiosity.

Like, all right, as a salesperson, I’ve got four people that I need to follow up with tomorrow and I need, you know, for business development. And if six people I want to follow up with the next day and seven people I need to follow up with the next day. And I have this list and they’re on a task list and sometimes I get to them and sometimes I don’t. And so as I’ve kind of matured my skills as it relates to AI. I thought, wouldn’t it be cool if at midnight every night, it knew which leads or which contacts I wanted to follow up with. And so I’ve built this workflow. It’s pretty simple and it will be more elaborate as I find time. But basically, what it does is it says, “All right, Matt needs to follow up with Todd.” So, pull the last six emails either from or to Todd. Pull the last AI note-taker summary from our last Zoom call and just create a follow-up email for me and put it in my draft email. So, when I woke up the next morning, I got five emails in my draft that were about to the 90 to 95% point. And so I do make some little tweaks because it’s not perfect. It doesn’t know exactly what I’m thinking, but I give it enough context that it, you know, that it’s really close.

I’ve also trained it on my tone of voice. It knows how I write it. I challenge you to find all that sounds like AI. Not anymore. Right. So, this is the stuff that adds value. Right? And so you string enough of these projects together and now clients are saying, “Can you do that for my sales team?” Right. “Can you do that for my marketing team?” So, all my little projects are how I build social media and how I build LinkedIn posts. How I refine my tone of voice so that it really and truly sounds like me. In fact, it’s actually better than me now because you’re actually getting the value of the LLM. And you know, that’s just helping me be super productive.

Todd Merrill:

Okay. Yeah. Neat. I don’t know if you want to, you have to share some with us on screen.

Matt Oess:

So, would it be helpful? You know, I did that demo with ElevenLabs, which kind of shows the language capability of some of those things. Would that be interesting to look at?

Todd Merrill:

Sure. I mean, whatever you have. So, to me, you know, software people are looking at this going. So what? We’ve had those tools forever. Right? Like I’ve been doing this stuff. We did a Chick-fil-A drive-through language model, you know, trying to pull out named entities that were on the menu. You know, I want to chicken nugget meals, but the aha for me. So, there’s a lot of software people going, like what is the big deal here? And the big deal is that you know, you don’t have to hire two or three or four software people and a project manager and make a product to solve the problem.

Now these things are all readily available. And then it’s an 80, 90% solution quickly to tackle some of these automation problems you’ve had all along that just weren’t worth spending the money to fix. Now there’s a bazillion cool little tools you can string together and get most of the way there. You know, is it going to write your emails for you? Not all of it. And you do want to have that final human check. But boy, you know, you don’t have to spend an hour spinning up coffee and thinking about what you’re going to write in this email. And then the next one, the next one, it’s all just kind of there. And you go, oh, yeah, that’s right. And I know this other bullet point needs to go in there. Yeah. Okay. That’s interesting.

Matt Oess:

This is just a draft, right? But I’m trying to think about the maturity of AI. And I think your audience might be interested in this. So, most of us have played with this, which was what we call zero-shot generative AI. And this is like when you use ChatGPT, hey, you know, refine this email for me. You know, write a blog post. You know, help me write this code. So, you give it an instruction and it sort of comes back with the answer to that instruction. Right. So, everyone has played with this.

2025 is the year of what they’re calling Agentic AI, where you have these agents that maybe you can just buy or that you develop. They just take on a complex task for you. So, for example, Synthesia is a really cool way, where you basically could use ChatGPT to write a script and then it creates a video with an avatar of somebody delivering a presentation or a sort of podcast type of thing, etc. But whether it’s AI copilot for code development or free pick, there are just these agents that will go do a task for you. And more and more companies are starting to get to the point where these things are pretty good and you can buy that for 30 or 50 or 100 bucks a month or kind of in order to do those tasks.

So, this is make.com and this is an automation platform. It used to just automate things, but now it automates things and lets you access all of this AI goodness along the way. And so if you’re not a coder, but you want to start to play with this stuff and understand this stuff, I highly, highly recommend this. And maybe we’ll drop a link or two for your audience if they want to investigate some of these things.

But what’s really cool about the things that I’m working on is what we call this multi-agent where you might have one agent that’s a director and you tell it to do something and then it makes a decision on how to go and execute that. And then it’ll just go off and do that autonomously and then bring you back the answer. So, this one that I’m working on is kind of my version of Siri, right? So, there’s an input front end. Right now, I’m using this app called Telegram, but it could be WhatsApp or it could be Slack or it could actually be Siri. And you just use your voice to tell it what you want. It does kind of a transcription and kind of understands it. But this guy right here is a ChatGPT module. And what it does is it sort of categorizes what I’m asking it to do. So, I could say, “Hey, book me an appointment with Todd for next Tuesday at 9 o’clock. We’re going to talk about X, Y, Z.” And it would know that is a scheduled thing. It’s going to go to my calendar. It’s going to build that thing and it’s going to report back that it did it. Or maybe I’ll ask it, tell me what’s on my schedule for tomorrow. Or maybe I’ll say, I need you to add a task in ClickUp because ClickUp is my task manager, my project management. And it’ll just go and do that. But I’ve also sort of trained this in my life. And so if I ask it something that it should know about me, it will go to what’s called this rag vector database and it will pull information that I’ve trained the model on.

You guys have heard this, training the app, training the model on all these things that I care about. And so this is how you take this. The biggest complaint about generative AI is it doesn’t sound like me. It doesn’t really know me. It’s just too generic. It sounds generic. Well, now you can see that we’re taking the steps to make this thing bespoke to you, your tone of voice, your business, your solutions, your pricing, etc. And so this has been really fun to play with and I’m really enjoying it.

Todd Merrill:

Yeah. So, for me, this is fun, right? So, you’re able to access these technologies that we’ve been sitting on for a while, maybe a little bit more advanced or more polished for sure. We’ve had vector databases for a while.

Matt Oess:

Absolutely

Todd Merrill:

Right? It’s just been really hard and very expensive to load them up. And you got to clean the data and then you got to train it. And then I think the big aha unlock was seeing you go through this and you’re an old coder, right? So, you got all these little bubbles, and then I think at the end of this, at one point you said, and now I’m coding. Yeah.

Matt Oess:

Yeah

Todd Merrill:

Like in the software brain, we think in terms of rapid prototyping, right? And then some problems aren’t worth spending money to fix, right? Cost benefit. So this lowers the bar to key ahas for me that you helped me unlock, right? So, what’s the big deal? We’ve been able to do this forever, right? And for me, it’s very low cost because I can, hey, just knock it out code. And we’ve been able to do a lot of these things forever, but now you don’t need a coder or team of coders or an elite anything, right?

Matt Oess:

Yeah

Todd Merrill:

 It’s kind of like what Zach, Zapier did for all these little SaaS apps where you can kind of glue them together and one of them is WordPress. And then another one’s like Stripe for payments. And you can kind of glue these little SaaS apps. This is the next logical evolution.

Matt Oess:

Right

Todd Merrill:

There’s a business layer in there. You can make decisions and try to figure out what they’re saying and route it and do all kinds of cool stuff.

Matt Oess:

Yeah. So Zapier definitely got here first, right? And I would say that, you know, make.com, sort of, is the up-and-coming competitor that in a lot of ways is sort of, you know, much better. And there’s a new one out there called n8n that I’m really excited about and I’m just starting to play with, but I think that’s the beauty of this whole thing is there’s just always something new. There’s always something new and there’s always something to learn. And it’s really challenging for me to make sure that I don’t have shiny object syndrome the minute that something gets announced or whatever because there are some things that I just want to get done. So, you mentioned coding and, you know, I literally haven’t written code for 30 years.

Back in the day, I worked in this advanced development lab for General Motors and we incubated in the early 90s some of the stuff that’s in cars today.

So, XM satellite radio. At the time, we called it digital audio broadcast, and it was way too expensive to think about putting up satellites. And so we did it terrestrially, but all that technology just got ported to satellites and that’s what everybody uses today. But I worked on OnStar. I worked on turn-by-turn navigation, and, you know, it was just a really extraordinary time where I’m writing assembly language code and C++, and I’m basically creating, you know, prototypes of hardware and soldering down chips and things like that. And then like I said, I turned in my engineering card and 30 years later, I’m writing code again. And it’s only because of AI that I could possibly do this because it wouldn’t be C++, it’d be TypeScript or it’d be React or it’d be, you know, JSON or whatever. And so what AI allows me to do is to write an app.

So, you know, hopefully in the next couple of weeks will be the world premiere of the app that I’m working on. And with the help of AI, you know, I’m able to do it. And as you say, I don’t need a team of 10 coders. Now, I hope that it’s wildly successful so that I need to add 10 coders, but today, it’s just been awesome to just play with this stuff and build something that I’m proud of.

Todd Merrill:

See, and that’s the key phrase, is you get to market very quickly. You could test product market fit, which is all the things we preach for early-stage startups, and then figure it out. And then when it is time to scale up, yes, these tools can help you start, but they’re not going to scale. They’re only scaled so far. And then you hire the engineers and then therefore now the engineers get to work on not like, you know, can I debug my C++, you know, code driver. We’re working on concepts that are big and meaty and more meaningful and more into the business domain. And then, so I don’t know this is kind of maybe let’s twist into like a little bit of a philosophical discussion on, you know, as you get more and more experience, you know, we all come from different parts of the world, sales or technology or what have you, finance, at the very top of the experience, strategic experience C level. Yeah. It seems like there’s a lot of people that have a lot of overlap, like you’ve had to kind of absorb parts of these other domains.

You know, AI can kind of help you get through the learning curve for a lot of these domains. But ultimately, you have to understand why you’re doing things and what the fundamental strategic problem is that you’re trying to solve or else you’re kind of wasting resources and time. Right?

Matt Oess:

Yeah. Agree. Agree.

Todd Merrill:

Where do you think all this is going? I was kind of joking, you and I were bitter rivals back in the telecom days, you know, Nortel and Ciena. I think actually, I’m not sure who bought who, but I think Nortel and Ciena merged at some point, you know, so it all came together.

Matt Oess:

Yeah

Todd Merrill:

But, you know, back in those days, it was the same kind of feel, right? It was go, go, go brand new technology. And oh my God, how are we going to get this done? Lots and lots and lots of money, gold rush until it wasn’t, you know, do you feel like this AI cycle is like that or different from that last technology, maybe not the last technology cycle, but a big one for sure that we had in the past?

Matt Oess:

I guess in some ways, I kind of, I resonate with the kind of what you’re saying and what we went through and when the bubble happened and the bubble burst in the early 2000s, you know, cause I was in the optical long haul business. I think, you know, Ciena did some long haul, but very much metro and locally.

Todd Merrill:

Yeah. I wrote the OC192 optical module first, which yeah. So yeah.

Matt Oess:

So, you know, at the time everyone is like, oh, we’re going to need all this data. You know, all this stuff is going to … you were going to need this for video and all these things. And we were in search of this killer app. And so they laid out like 396, you know, fiber cables all over the place. And each one of those fibers could have a whole bunch of OC192, 10 gigabit data streams going bidirectionally in each fiber. And then you have 196 fibers and then you have lots and lots of these bundles of 150 fibers. And so it’s sort of like, it was just sort of a race to this future that we didn’t know was going to happen. Right? And at the end of the day, it was like, oh, there wasn’t as much traffic as we thought there would be. Right? And so the bubble burst and all of the valuations just went crazy and went into the tank.

So, in some ways, it does feel a little bit of like what is the upper limit on the number of big companies that have to spend billions and billions and $500 billion to build out server farms, to train models? Like, at what point do we have enough people doing that? That like each, if you think about the top 10 companies in each industry; pharmaceutical companies, the construction companies, the tech companies, like the whatever, you know, are they all going to deploy all these things? And so you wonder where the upper limit is on the number of chips that we’re actually going to sell right before it’s like, okay, we probably have enough of that.

I think if that were the end, this might feel very much like 2000. But I think this is the beginning because of the application side of this. And so I feel like that is the piece that people are just now starting to get their arms around. Like this, you know, the stuff that I’ve been playing with, you see the value that it actually creates. And now all of a sudden, like you can actually see where this is going to go on for a long time. It may not mean for semiconductors, but like the, you know, Palantirs, these people that are trying to figure out how to defend us against 180 drones that are coming at us at the same time, or 5000 drones that are coming at the same time. Like, there are apps that are going to leverage this stuff. And it’s not just defense stuff. And it’s not just pharmaceutical stuff and it’s not cancer research. It’s our everyday task.

Todd Merrill:

Yeah. And then, you know, in the same way that the fiber optic telecom boom happened, it didn’t go away, right? We just found other more mature applications for it. You see a lot of Ciena or Nortel gear in data centers now, right? So, who’d have thought, you know, and then I think the same thing is happening. You’re seeing this big gold rush, who can build all these data centers with cooling and adequate electricity, cheap electricity in and around the United States, you know, it’s going to power a bazillion small use cases, and it’s going to make us all a lot more productive in the world.

Matt Oess:

Yeah

Todd Merrill:

But then we’re going to have to start working on bigger problems and what are those interesting, bigger problems that we can now address that we couldn’t before?

Matt Oess:

Yeah. And the last couple of weeks, I’ve been thinking about what the future of some of these companies looks like. Like, if you think about some of the big powerhouses in the last 20 years, sort of the salesforce.com and the big ERP companies, and Oracle and SAP, etc., etc. They’ve always kind of been the database, but they’ve also been sort of the application side for your ERP or for your sales, etc. And, you know, if you’re Salesforce, and you’re going to charge $1,500 per salesperson or per user per year, I just wonder if that model is going to change if SaaS changes. So, one of the things that I’ve started to see is, and I just wonder, does Salesforce become the data repository, right? For all the CRM stuff and all your deals and all that? And is that the right place for that to live? Right?

But do they keep up on the application side of this, right? Are they the ones that actually create the sales agents and the marketing agents and the drip campaigns and the content? Or are there smaller companies that start to do that piece of it because they can innovate faster. So, for example, if you’re Salesforce, and you’re trying to figure out how to create an agent for a salesperson to be more productive, are you going to be able to take every single one of your clients and train the AI model to fully understand their solution set, their context, all of their emails, like everything that you’re going to need in order to like, let AI create and do the intelligent things that you want it to do. I don’t know if they’re going to be able to do that. And I don’t know if they’re motivated to do that. Because if they do that, can they still charge $1,500 a year or $700 a year? Or do they sort of get relegated to, you know, to being the database? The other thing that … anyways…

Todd Merrill:

… yeah, yeah. no, I have thought on that. So, you know, I have discussions with clients all the time on this and then it’s like, this is just APIs again. Well, but it’s not, right? And then, oh, my God, SaaS is going to change. Yeah, well, yeah, it is, but it’s not. It’s an evolution. And then so, the way I think about it is we have to up our game and how those things talk together. And that’s really all technology has ever been, is taking ever more sophisticated Legos and snapping them together in creative ways as fast as possible to get the right answer to make people go faster to solve problems. So, what I tell people is start thinking, you know, there’s going to be agents out there in the world. Your clients are going to all have their agents.

Everybody is building agents right now. Well, get ahead of that. And then much the same way, we always used to tell people, hey, VC play is go get some money, get your SaaS out, and then you got to have an open API. You give tokens to people and then make it easy to integrate. Well, same holds true. It’s just a different way to do it now. If you’re Salesforce and you’re lead management, right, you’re going to have to interoperate with lots and lots and lots of agents, with lots of smart people doing lots of creative things. So, we have no idea where that’s going to go. But they’re all going to need lead management at some point in some creative ways and like smash this all together. Be agent-friendly. And then, you know, be really, really good at that one domain. And then provide your software in that way. It’s not a UI anymore. It’s not an API anymore. It’s an agent intent processing node or, you know, something we need to name. But you start to see those protocols come out where do you take your deep thinking, deep, you know, meaningful piece of software or business intelligence or data pool? You’re not going to get away from data. Clean data is the lifeblood of ML at the bottom of it. Right?

So, how do you bring that data pool or that business intelligence or that deep thinking or that other agent or the machine learning or whatever we want to call it to bear into as many different problems, domains as possible? And the answer is you have to figure out a way to plug into somebody else’s agent on a protocol. And I think you’re in, you know, I was looking at your blocks, the maturity model, I was kind of giggling because it’s like, that’s right there with three moving to four. You know, that’s how you do it.

Matt Oess:

Yeah. That three moving to four, like, I guess I really didn’t cover that in the slide. But I think eventually, people will start to embrace these multi-agent solutions. And they’re going to actually, like, I think that some people, maybe big companies, maybe pulled the trigger too quickly. Right? And they did these million-dollar science experiments that didn’t create an ROI. Right? Because the application layer wasn’t there yet. Right? But I think that as people start to create these custom multi-agent scenarios and they learn what they need to learn on a corporate level, and you start to get some at-bats, and you start to see the value of this, then I think it leads them to a whole bunch of innovation. Where they can actually start to leverage this stuff to innovate, not just in their workflow, or their processes, or their backend, or their accounting, or whatever, but to actually help them innovate their products. Right?

And then I think that the next step is they’re building their own custom models, and they’re training their own models. And then it’s almost like an AI-first sort of thinking. And that’s where I think we’re on our way. I think that’s probably nine to 12 months away.

People need to sort of get past just zero-shot GPT. They need to play with some of the tools. And so, a lot of the work that we do, there’s a small team of experts at TechCXO that are in this AI practice, and it’s growing really quickly. But some of the applications that we’ve already deployed for clients are just really unbelievable. And clients are, you know, they’ll meet with us for these things, we call 60-minute discovery sessions. Right?

Well, we’ll just sort of do some show-and-tell of some of the cool things that we’ve been doing, and talk about projects that we’ve done. And then we just ask them what’s getting in the way. And if you think the promise of AI could do something that would really create a step function change in your company, let’s explore some of the things that might be. And then let’s do a proof of concept, like for two weeks or a month. And then let’s just stand it up and see if we can extract some value without going into a crazy investment. And then once you start to say, oh, there’s a thing here, then kind of, as you said, you start to bring in more engineers and you harden the code and you put all the error, you know, handling and all the things that you need in order to scale it.

So, if you look at that make.com situation that you need in order to do rapid prototyping, I don’t think we’re too far in the future where you just export that thing, you drop it into a code assistant, and it just creates it for you and then off you go.

Todd Merrill:

Yeah. Yeah. I’ve seen that movie before. It got burned. Yeah. I would love that. That would be an amazing future. And, you know, we go through these cycles and then, you know, I kind of go, oh God, here we go again, you know, with the no code. But every cycle of this no code or code builder or report generator, you know, 4GL, going back in history, gets a little better, a little better, a little better, a little better, a little bit less of pain in the butt.

Matt Oess:

Yeah

Todd Merrill:

Ultimately, you’re going to have to have smart people working on complex things, but hopefully, that comes later when you have more proof of concept, more certainty, less risk in a project. And, it’s a sure bet, you know, and I think that’s the key takeaway for a lot of these projects, is as we get better and better technology, how can you make more stakes quicker with less investment, and then just kind of test very quickly and get a whole lot better result for the whole world.

Well, Matt, it’s about getting out there in the world and talking to people and making things happen. You got any great travel stories you want to share with us?

Matt Oess:

Travel stories. You know, I just happened to pick up literally a file, like it was full of PowerPoint, you know, slides. And I thought, man, when was the last time I printed out a PowerPoint slide?

But anyway, I got a call out of the blue and it was from Ernst & Young.

And this is 2018, 2019, when I was still traveling a lot. And the call was from Ernst & Young. And I was like, is this a scam? Like, Ernst & Young doesn’t call me. Like, how did they get my name, etc.? And so it really took me a while to sort of get comfortable with the conversation, but they were like, you know, you came highly recommended. We need a strategy facilitator. Can’t tell you who the client is until we get an NDA in place, but are you available to go to India and conduct this two-day strategy session? And the whole time, I’m thinking, seems like Ernst & Young would have strategy facilitators. Like, why would they want me for this?

And so anyways, you know, I continued the conversation and figured out that it was legit. Signed the NDA, turned out to be Sony Network Pictures Incorporated, which is like a massive, you know, kind of cable company in India. Right? And so, you know, they wanted help with a five-year strategy to go from a billion to 5 billion in five years. And so I met with the 13, the gig was, you know, I met with the 13 C-suite executives and I worked them through a two-and-a-half day session to come up with the new vision and the new plan, etc. But the interesting part of the story is, you know, can you come and facilitate this and can you be in India in nine days? And I thought I don’t have a visa. I don’t even know what it means to go to India. Certainly, I have never been there. Don’t know what I’m doing. And they figured out all the things that they needed to do. So, I got on a plane but didn’t know where I was going. I landed. There was somebody who picked me up in a car, you know.

I ended up in Ernst and Young’s offices there, you know. I think kind of scoping me out and also preparing me for the people I was going to work with. And then we fly to this place called Udaipur, which is one of the regions. One of the cities and one of the regions centrally. The hotel used to be a palace. And so the palace is like every floor, every wall inside and out is made of marble. And so you walk through this metal detector, you pull up in sort of the black, you know, SUV and you walk through the metal detector and there are guys with automatic weapons standing there. And you’re like, where am I? And it’s a hotel that happens to be one of the top 10 hotels in the world. Right?

And so it’s like a football field in this direction, a football field in this direction, a football field in this direction of used to be a palace, and now it’s like this, you know, $8,000, $9,000 a night hotel. And this is where I actually got to do the facilitation of this thing with these executives. So, I’ll just never, ever forget it.

Todd Merrill:

Wow! That’s awesome.

Matt Oess:

Literally, there are peacocks walking around in these gardens. This place was so extraordinary.

Todd Merrill:

How cool is that? That is awesome. Wow! Well, how can people follow up with you after this on the internet, if they want to get in touch with you about some of these things?

Matt Oess:

Oh, we could splash up a link for them. LinkedIn is a great way to get me DMs. Of course, matt.os at techCXO.com. We can show them that, in the credits or whatever, but I would love to talk to you about anything that we talked about today.

Obviously, AI is a big topic for me, but you know, anything regarding coaching, etc., or if you just have a need, we have an incredible bunch of partners at TechCXO and they’re truly experts at what they do. And I think that especially in 2025, there’s just a bunch of uncertainty and we have to make some really good steps and we can’t have missteps because I think there’s just a lot at stake. And so if you’re a startup that’s getting funded or trying to get funded, or you’re the company that needs to do some kind of a transformation or you just need to level up, you know, I just don’t think this is a year for amateur hour. And so, you know, if I can connect you with some of my partners to help you in the C-suite to try to further your efforts and give you the best chance of success possible. I’d love to be a service.

Todd Merrill:

Great. Well, thanks, Matt. It’s been great having you.

Matt Oess:

It was my pleasure. I enjoyed it. Thanks for having me.

Todd Merrill:

Bye for now.

Matt Oess:

Thank you.