Tech Stack Reimagined: The Reality of Building an AI-Native CRM

[00:06] **Introduction** Chris Carolan: Good morning, good afternoon, LinkedIn friends, Value-First Nation, welcome to another episode of Tech Stack Reimagined. Uh today talking about the reality of a vibe coding your own CRM. Uh how are you doing, Nico?

[00:31] **How is Nico Doing?** Nico Lafakis: Doing good. Doing really good. Um you know, this uh this topic in particular just coming at a quite a pivotal time. Um But yeah, I I'm doing very well. Uh especially with the nice little data leak of Mythos. happen. Uh happened on Friday. So, um that's nice. I'm sure that I don't know when we're going to get to see that. But I'm just glad uh about the description that was given. And the fact that Dario is considering it to be like it's new, right? So if you guys think about the model releases, obviously Open AI is like the worst at naming. So it's really difficult to tell sort of whether like four and and five are that different or like what's different about four to five in terms of the architecture or like is it just an upgraded version, right? Whereas with Claude, you really do feel and understand the difference between Haiku, Sonet, and Opus. There's a very, very clear distinction, especially when it comes to like coding or even when it comes to complex reasoning work. You can very easily see the difference between all three models. If you work with API, you very much understand the capabilities of all three and like what the outputs of all three look like.

[04:21] **The Leak from Anthropic** Nico Lafakis: The having been said, excuse me. The leak from Anthropic is about their newest model and the newest model has a new name and the name is Mythos. And that means that it's very different from everything that came before it because it's a new name, it's not just Opus for, you know, five and Sonet five or whatever, which is what I was expecting. I was certainly expecting just the fifth, you know, the newest generation of those models. And that may still be the case. We may not see Mythos for quite some time. A essentially, the reaction to it was oddly. I'm not sure like I don't know I don't know the rationale behind these things. It's not making any sense. There's really no pattern to it outside of maybe just shorting certain market sectors for whatever for one reason or another. Um because, you know, we people, I guess like the market woke up to Claude code, right? Like what a month ago and immediately shorted software stocks all together, right? Specifically CRM stocks. Just kind of kind of odd to me. Then the Mythos uh leak happens, which by the way, if you were thinking that this was like somebody hacked into Anthropic and managed to get this data, nope. Anthropic even admitted it was human error. Some some person forgot to lock the door behind them when they left after doing certain blog updates.

[07:03] **The Meathos Leak Was Going to Be Announced Eventually** Chris Carolan: Yeah, like draft email like or not draft email. Draft blog article. Yep. Right.

Nico Lafakis: So that's the thing, it's like it was going to get announced sooner or later. Like pretty fairly soon, right? Um so it is, you know, it's really just like a pre-announcement if anything. But it's the commentary that Dario threw behind it by saying it's so so much more advanced that they have to give it to security companies like like cyber security companies first before they can even remotely distribute it elsewhere, right? Um and so all of a sudden, cyber security stocks tanked. And I'm really like that's what's blowing like it's I just don't understand it. I I don't understand the mentality of the financial world where you have somebody saying we're going to impart the smartest thing ever created into security into like digital security. You would think that that would make any digital security company now worth just an insane amount more than it would have been otherwise because you now know that their level of security is about to go through the roof, right? But it went the other way. I kind of it doesn't make any sense but yeah, in terms of overall, you know, overall uh doing pretty good. How much stuff?

[08:31] **Chris Reacts to the Financial Reactions** Chris Carolan: I'm doing fantastic because this is like it's clear as day what's happening and like yeah, I've been working on these frameworks for like since December of 24. But like living through the dysfunction of software and SAS inside of organizations since like 2010, right? Like and all of this, like all of the everything that did not make sense to me like over 15 years is now getting solved over the space of a few months. like including markets, right? Like Yep. if like the transparency of what's happening like Scott stocks are tanking because it's not based on actual value being created, it's based on are we certain and do we feel good about the direction of this company? Uh in any given space, that's why like most investors or most like investment experts that I've heard as far as like if you just want to make sure you're going to get some return, like index, index funds, like go there, it just it's the market as a whole, you don't have to worry about which company because they're so fragile. Like when they can move like this. Like once uncertainty leaks in, really that's all it takes. And and when your confidence is based on feeling, like it doesn't take much to be like, oh shit, we better rethink this one, folks. And that's SAS. Like that's Right. that's what SAS is. SAS is about puffing yourself up so that you can get a bunch of venture capital to hopefully a unicorn. And and like the data behind whether or not that can actually happen is is not good. The chances are not good, it's a lottery ticket. And while you know, that's why I love listening to Moonshots. Obviously, it's people that live in that space can also paint some optimism around what's going on. They don't sugarcoat it. Uh I it's hard to I could see why it's hard to relate to what they're saying. Like just go build your own own own company like it's like, you know, people can do that, which they can now. capability wise with AI, but it's still like mindset of believing that any mere mortal human can do it and that's why we're on shows like this to show you how you can. But like the opportunity is there and this world has created private capital that leads to innovation and like it's has a lot to do with why we are where we are. But when you can see what's happening with the markets and the reactions and then people try to tie that to either the reasons why or defending that it'll bounce back, it always does. Like you're not paying attention then. Right. And this leads directly to this conversation happening on LinkedIn right now and it's funny because I think Hubspot as a company is way past like when they say like agentic customer platform, that's so far away from like the CRM like conversation and like just managing sources of data, right? Like if you're using a contact record like and like building your system around that being a a source of truth without any of the context involved that actually makes it a source of truth. Like you're just doing it wrong. Like and that's what's behind all these posts and and what led to this today um or what probably coincided with what we were naturally going to going to talk about. I'm going to share this post and there's like there's been like 100 of them in the past couple months, lot like this one, right? So everybody taking offense to just vibe code to a full Hubspot replacement in a weekend. Uh LinkedIn right now is I replace my CRM with AI agents. No code RevUp stack in three days. Goodbye Hubspot. Hello Vibes. Uh meanwhile in the real world, your vibe coded CRM doesn't stand, you should probably like move the quotes to CRM instead of vibe coded. Um Right. doesn't understand lifecycle stages, override good data with garbage, has zero governance breaks. The second sales rep uh goes off script. The second a sales rep goes off script and absolutely folds under attribution or reporting pressure. Now I'm going to I continue reading uh but commentary for each of these sections will be on this other tab uh like how to how to think about this. Um but the idea is like all of these true for every Hubspot implementation ever whether it was vibe coded or not. Mm-hmm. Right? Like most Hubspot implementations suffer from these five things. Like specifically and it's more than 50% of them, right? So here's why these things fail. So he just makes that case, right? No data model, discipline, no guard rails because you get bad data at scale. I build all that, well guess what, most Hubspot partners skip all that shit too. Most Hubspot implementations that are just onboarding sprints for 90 days miss all of this. And that's why people are coming back saying, I'm on my second or third partner, just gave me some damn help to do Hubspot right. Like and they're asking for it in different ways, like help us get the data model right, help us structure it. Help us with architecture. Like people are starting to understand what it takes to get this stuff right and those same people have AI in their hands. So maybe not look for the third third partner. Right. To maybe understand like okay, what's next? If AI told me to to ask for data architecture help, then maybe I can just ask AI to help build that out, right? And so what we're going to talk about today on show, definitely a lot of show is we've obviously been talking about this uh for a long time and for like transparent reference, we're both pretty high level consultants and coaches that know what the hell we're doing and talking about when it comes to CRM and even Nico who's who's like at that level, this was not like a one or two day job, right? Like and we've we've started talking about this a little bit. It's kind of like like anybody can pick up Street Fighter and win a win a couple matches after a couple hours of playing. Right? But if you go to like the arcade that has a line, line of folks or you go to like Nationals, you're going to get smoked. Right? That's what AI is starting to feel like. Easy to pick up, but as soon as you want to start handling super complex use cases, right? That it takes it takes some learning, takes some skill. Uh, right? So before uh and shout out to to Bill Barless here who like if you've got these symptoms, they don't sound like a tooling issue at all. Uh and my opinion nails it in terms of what we're going to hit here and then we'll we'll I'll unleash Nico uh with his CRM. Right. So uh a little decision framework. I'm going to put this in the chat, which by the way, this was built like an hour ago after getting ready for the show and please keep that in mind like if you prevent people from vibe coding shit, like it prevents this kind of capability and you can't prevent it at the end of the day. So built by neither question matters until you understand the traps. The false binary uh is it vibe coding or is it, you know, impossible to do that, neither is correct. You can absolutely you absolutely can build powerful custom AI native systems, but it takes real architecture, real security protocols and real effort. The technology you choose matters far less than the thinking you bring to it. The same traps that plague enterprise CRM implementations will plague your custom build if you do not design around them intentionally.

[25:27] **Twelve Traps** Chris Carolan: So 12 traps, we'll go through a few of them. The leads trap. Treating people as objects through process rather than humans on a journey. If you build custom CRMs recreate this trap with cleaner UI. If your data model starts with leads and contacts as separate objects, you've already encoded dehumanization into your architecture. No amount of AI polish will fix a broken premise. If you buy enterprise platforms reinforce this with MQL SQL pipelines, lead scoring and lifecycle stages that describe what you want from people, not what they are experiencing. So the question to ask, does your data model describe people as humans on a journey or as objects in a pipeline? Uh let's look for another good one here, some operations traps. The measurement trap. measuring what is easy instead of what matters. If you build custom dashboards make it tempting to measure everything. More data points do not equal better decisions. If your custom CRM has 47 charts on the homepage, you are in this trap. Uh if you buy platform ship with default reports optimized for vanity metrics, the out of the box dashboard tracks activity volume not value created. So ask this question, can you articulate the three metrics that actually indicate whether you are creating value for the people you serve? Uh the SAS trap. over customizing software to match broken processes. If you build, this is the trap vibecoders think they are escaping, but building custom software to replicate your broken processes is the same trap with a higher engineering bill. The process needs to change first. And if you buy every custom object, custom workflow and integration that exists to make the platform match how you currently work is a symptom. Adapt process to platform not platform to process. Uh are you designing for how work should flow or encoding how it currently stumbles? And then let me look for my favorite one. Uh the managed services trap. outsourcing thinking, not just execution. If you build hiring an AI agent framework or no code agency to build your custom CRM is this trap wearing a different hat. If you cannot explain your own data model, you do not own your system. And if you buy platform agencies manage your instance, you pay monthly for someone else to understand your own business processes. When they leave, the knowledge walks out with them. or on the converse, you tell them that somehow they've got to create value for you inside of three weeks and we we're not allowed to have discovery conversations to learn how your business works and you just we're just setting up everyone up for failure. So if your team can't explain your system architecture to a new hire without calling a vendor, you you got to solve for that. So 12 traps here, a little you know, a little quick test. Like can you draw your data model on a whiteboard inside of five minutes? If not, uh might be too complex or you don't know enough about it. Uh, can you explain who sees what and why? Can you describe how value flows through your system? And can you change it without fear? This this might be the biggest benefit to what we're talking about today, right? And the reason it is so powerful to have AI in the hands of folks that are in the shit every day. Like they're the ones that got to get worked on, they got to create value no matter what. And when they're unlocked to do this with AI as like the safety net to be like, hey, let's make it look like this real quick. Oh, it's not working. Okay, let's go back to the old thing, we're cool. Let's you know, so much of the SAS industry and tech debt, right? Our changes that get deferred because we might break something because we built a bunch of workarounds in the first place and now we don't know how anything works. And this power in the hands of the people in process has been so powerful to watch and understand that it's the immeasurable part of why if you're telling people that they're not allowed to solve their own problems, it's just not you're not going to win that argument. Right, as soon as they figure out that they can solve their own problems, anybody suggesting they can't is out out. Like either out from like a we're done working with you or we might be stuck in this contract but I don't trust you at all anymore and we might as well just count down the days until until we move on. Um so any of that resonate, sir?

[35:28] **Nico Finds Himself on the Other Side of the Wall** Nico Lafakis: Um, a lot of it does. And it's really interesting to me because um, I I find myself I find myself on the other side of the wall. Um and I I find myself there on purpose. So what I mean by that is uh to your point, uh I'm definitely somebody who has some background experience in working on the CRM both from client side and from admin side. And because of that, yeah, I definitely know the pitfalls both of platform and also of organization. The number one thing that I've noticed in all the time of servicing clients and whatnot is just that that factor of hey, it's really difficult to onboard people. It's really difficult to get them to understand the platform. We got to do trainings, we have to spend, you know, two hour sessions. Hopefully everybody's paying attention. Hopefully everybody remembers everything. No one's really going back to the training material. So in other words, when it when it came to doing that to attempting to get a business to change its process for the better as opposed to having to adopt some software platform to make it better. I think that that's been a negative in the past because the software couldn't solve the the problem, the process problem. I don't think that's the case anymore. I think the software now solves the process issue. And I think the software now makes up for the human inaccuracy. And it makes up for the human laziness and it makes up for uh the human uh I don't know, fallacies that go on. Um it all depends on how you built that platform. Okay. And so to the you know, look, not everybody is we we've talked about this. not everybody is like even at our level, which is not bragging in any way whatsoever, just saying like as much stuff as we know. Um, I agree with that post when it comes to somebody who's just stumbled onto Claude, just stumbled into vibe coding and you know, from maybe a not CRM technical perspective uh is trying to do it. Okay, then I agree because again, you don't really have you don't necessarily have the background knowledge that will keep you from making the same systemic mistakes. So you'll end up building another CRM platform that will almost inevitably inherit the same problems that other platforms have because you don't necessarily know what to avoid. You don't know what you don't know, right? Um in in my case, I think it's very different. I think in our case it's very different because we build things differently. We we understand that software has to evolve and that it's not human first anymore, it's agent first now. And so if you're building a platform that's agent first, it is amazingly human intuitive.. I know how odd that might sound, but the the reason and it's funny because I excuse me, had a conversation with my brother yesterday and kind of came to the same conclusion uh when it comes to, you know, labor getting replaced and what types of labor are going to get replaced and Dario just sort of put together another list of like top jobs, not him, but they did another study and put together a list of like top jobs that'll be effective affected by it and all that. And of course, you know, manual labor and and uh city workers and stuff like that, electricians and whatnot still way at the top of that list of like, yeah, not going to get replaced for a while, right? Which also has to do with uh robotics and things more than anything else. But when you ask the AI system why it will take so long, it's not the dexterity of the hand. It's not the physical capability of the machine. It's not the like inability to see certain things. It's this strange ability to adapt to random situation. And when I dug a little bit further and I'm like, well, what exactly are you talking about? Well, it talked about the same things that humans run into that they complain about all the time, which is to say that if you go to one city and you go to one city sewer and the way that one city sewer drain works in section AB, and you go to another city and you go to that city sewer and you go to that city drain and section AB of that city, they're not the same. They're done totally differently. They have a completely different system. If you go to the next city, even different from the other two cities. Go to the next city, different from the other three guys. Why? Oh, you got different different people working in different cities. They all have their own method of getting it done. So it's completely customized to the city. So of course you're not going to be able to get one robot to understand how all 50 states and all the cities in every state like how all of that stuff functions. No, it's it's going to take a minute, right? That's the only reason. It's not like, oh, the work is impossible, they couldn't possibly. No, they can. They definitely can. It's just that it's the job itself has become over complicated. And so I've watched this same thing sort of like yes. Looking back on the platform of like, huh, well, that seems what's happening with like software platforms, which is making them more and more complicated to the point where you have to reach out for extra help in order to understand the overtly complex system that you're having to deal with, right? Yeah. So, yeah, I mean at the end of the day if Yeah. to me, what you're saying makes perfect sense. Like Right. and this is the full circle, right? The SASpocalypse is actually the great simplification, something I'm going to see see if that starts to hit because basically and it's like what's been happening this year driven by Anthropic. When Anthropic dropped MCP in December of 24. They basically said, fuck all that. You guys don't get to decide how your APIs work and how all your systems work and because of that complexity, we got to build iPads and we got to have integrators from both sides. We can't just have integrators in the middle. We got to have people that know both systems because they're so dramatically different. We've got to figure out how they're going to talk to each other. It's like, nope, yeah, we're we're solving that like stat. Otherwise, AI, yeah, they can go, you know, read all of the, you know, API docs and all that, but it's every when every system like every city or every human is deciding, no, this is how we think it should work, right? And when the DNA of our software comes from like some very specific vertical use case and they're like, oh, there's a market for this. Let's build it for everybody. Yep. Right? Yeah. And and it and there's so that's why we have 17,000, you know, apps on the martech map, right? Because that's like a it either has come the from there or it goes the other way. It's like let's build our own like, you know, nonprofit specific version of CRMs. And then oh wait, but it's only based on our knowledge of nonprofits and not based on like a hundred nonprofits and how they work. So the next one's like, oh yeah, we need to do this. It can't work. Right? Um, so let's integrate the real shit into the right? And that's where when you treat these systems like if Hubspot is configured as records to process, replace Hubspot with Salesforce or NetSuite or any system even the way your business operates, right? When people operate like that, they will be replaced by AI who can definitely manage records to process at scale with faster and more consistency than any human ever has. Now when your system is designed to require, which it naturally is, and that's the difference like if you don't see this about the system, it means your humans are doing way more work than they should be because all the context management lands on them because they don't have a system that can help manage it at all, right? Right. So these are the humans that it's not in their job description to manage six different systems, but they know that's where all the data is to answer the question they just got from the customer. Right? And it's like the human context piece of that like sewer situation where a human can go and like, oh, I was expecting something but it looks just like this other time that I had to do it, but not exactly because that time was like this and this time was like that. And now immediately like our brains can can put that together. Like that's human context and judgment, right? That Right. we can build into these systems, but it requires a level of complete visibility to everything that's going on that you can get to, but you have to ask these systems to do that for you. And you that's where agents working with humans when you can build it out for that experience, it's the difference between like how how are you thinking about this, right? Are you trying to build a unified platform that goes across the business and like, you know, including humans? Are you cool with just, you know, building in like like point solutions everywhere, right? Narrow data wrappers, AI can easily replicate, super UI heavy, manual data entry, super focused on reports. And all the seat based stuff, right? That's that's gone, that's going away. And so much of this like is inflated cost of SAS, right? And we're seeing it and we're seeing it and it's like basically simplification means all this stuff is gone.

[46:21] **Nico's Perspective, and Reaching Different Levels** Nico Lafakis: And if you guys if you guys think that like so we're borrowing, so like what we do, you know, what Proven Labs does, right? Um what we do is we basically put you know, do that uh I mean throw that middle finger up to the SAS industry, right? Um, it's kind of the the purpose and the and the point of what we do and it's everything that um Chris was just talking about in terms of we solve that problem. We build that custom platform to solve the problem that you're having, right? And then you own the platform. You you take it, you do what you want with it. we're not we're not in the business of trying to tie you down with monthly payments, right? We're in the business of trying to solve problems. Make money by solving problems. That's last last time I heard when it came to commerce, right? That's the number one way to to make money is to solve problems, right? So, um that's what we're focused on doing. And to your point, I was just talking to to my wife about this this morning. Um, you know, the what's happening to SAS and this whole SAS apocalypse thing, right? You've got the release of Claude that tanked the, you know, CRM market and software market in general and now you have this Mythos leak and that's starting to tank, you know, uh the market around IT and all that kind of stuff. There's something that you guys have to really keep in mind. One, the way in which the great majority of corporations that are traded online, they are not PBCs, they are just C Corp, just regular corporations. They are legally beholden to make more money. It is their legal obligation to the stockholders to make more money. That's the only obligation they have. They don't have any sort of like ethical obligation at all. That's why companies do the most unethical things to make money. PBCs are different, public benefit corporations are different. They cannot exceed one for the other. They cannot exceed research for the sake of monetary uh gain and they cannot forsake monetary gain simply to to advance research. Um, works basically to keep a you basically take a nonprofit organization, make it profitable while still maintaining the capability of keeping the money from controlling the research and and the release of product. So that being the case, realistically what's happening is the value of what these companies have placed on these platforms is becoming normalized. The best way to look at this is uh Louis Vuitton. See the average Louis Vuitton bag, I think it's like minimum 1500 bucks. 15, 1600 bucks. Uh last I checked, maybe it's two grand now, who knows. Doesn't matter. If you actually research what the material cost and time and effort cost is for that bag. just in general. few hundred dollars. Maybe 500 bucks. So you are paying three times the cost of what it takes to make the bag because it has the name Louis Vuitton on it. That's it, right? If what is it that those companies like try to put the lockdown on the most? Knockoff brands. Well if Louis Vuitton sells for 1500, the Louis Vuitton knockoff sells for 500. Which is what you're supposed to sell it for and still make a profit. You're just not making an obscene profit, right? So the whole charge per the seat and the whole we're going to do a subscription thing. That's a way for companies, trust me because I know now to be able to play the odds just like a casino and say, well, odds are based on usage, yeah, we'll give you, you know, whatever it is, you got to pay for average company or whatever, 15, 20, 30 seats, right? And we know that even though you're paying for that, the amount of usage from those seats is so low that you are essentially maybe covering the costs for the actual cost for five or six people and you're just throwing the money away on everybody else. Right? Because how often, how often is a sales rep in platform? How much time are they spending in platform every day? Isn't the goal to get them to not spend time in platform? But you still have to buy a seat for them. But they're not sitting in platform all day. You see you see what's happening here folks? Do you understand where I'm going with this? Right? You're getting charged for things that aren't actually happening and you're paying for stuff. This is this is the clay game. all of these software companies have managed to do this 8020 game where you use maybe 20% of the platform and pay 80% for nothing. Right? So of the cost that you're you're paying to these guys, who knows how little you are using of it, right? A lot of these companies the way that they paywall their features, they'll paywall the most important feature behind a gate that holds, I don't know, whatever, 30 40 other features that would be useful for another company, right? But it's not useful for yours. But you have to pay the extra anyway because you need the two or three that that one gets. Instead of I don't know. talking to the company and saying like, hey, can we just get these two or three? And then what happens? Oh, no, we can't we can't do that. We we don't we don't have that. We have that as an offering. Really? Why not? Why don't you just you know exactly what the usage cost is for that thing. Why don't you just break that off and do it for like this particular customer. You own the business. No one said you had to offer that as an everyday product, right? But like even that doesn't happen with SAS. So yeah, I mean at the end of the day if to me, what you're saying makes perfect sense. Like Right. and this is the full circle, right? The SAS apocalypse is actually the great simplification, something I'm going to see see if that starts to hit because basically and it's like what's been happening this year driven by Anthropic. When Anthropic dropped MCP in December of 24, they basically said fuck all that. You guys don't get to decide how your APIs work and how all your systems work and because of that complexity, we got to build iPads and we got to have integrators from both sides. We can't just have integrators in the middle. We got to have people that know both systems because they're so dramatically different. We've got to figure out how they're going to talk to each other. It's like, nope, yeah, we're we're solving that like stat. Otherwise AI, yeah, they can go, you know, read all of the, you know, API docs and all that, but it's every when every system like every city or every human is deciding, no, this is how we think it should work, right? And when the DNA of our software comes from like some very specific vertical use case and they're like, oh, there's a market for this. Let's build it for everybody. Right? Yeah. And and it and there's so that's why we have 17,000, you know, apps on the martech map, right? Because that's like a it either has come the from there or it goes the other way. It's like let's build our own like, you know, nonprofit specific version of CRMs. And then oh wait, but it's only based on our knowledge of nonprofits and not based on like a hundred nonprofits and how they work. So the next one's like, oh yeah, we need to do this. It can't work. Right? Um, so let's integrate the real shit into the right? And that's where when you treat these systems like if Hubspot is configured as records to process, replace Hubspot with Salesforce or NetSuite or any system even the way your business operates, right? When people operate like that, they will be replaced by AI who can definitely manage records to process at scale with faster and more consistency than any human ever has. Now when your system is designed to require, which it naturally is, and that's the difference like if you don't see this about the system, it means your humans are doing way more work than they should be because all the context management lands on them because they don't have a system that can help manage it at all, right? Right. So these are the humans that it's not in their job description to manage six different systems, but they know that's where all the data is to answer the question they just got from the customer. Right? And it's like the human context piece of that like sewer situation where a human can go and like, oh, I was expecting something but it looks just like this other time that I had to do it, but not exactly because that time was like this and this time was like that. And now immediately like our brains can can put that together. Like that's human context and judgment, right? That Right. we can build into these systems, but it requires a level of complete visibility to everything that's going on that you can get to, but you have to ask these systems to do that for you. And you that's where agents working with humans when you can build it out for that experience, it's the difference between like how how are you thinking about this, right? Are you trying to build a unified platform that goes across the business and like, you know, including humans? Are you cool with just, you know, building in like like point solutions everywhere, right? Narrow data wrappers, AI can easily replicate, super UI heavy, manual data entry, super focused on reports. And all the seat based stuff, right? That's that's gone, that's going away. And so much of this like is inflated cost of SAS, right? And we're seeing it and we're seeing it and it's like basically simplification means all this stuff is gone.

[58:33] **Nico Previews His System** Nico Lafakis: All right. So, like to the to the point of like everything that we've been talking about and like the whole vibe coding your own CRM thing, um, I've done just that, right? I as you guys would expect. Um you know, we need a CRM for our business. I have options. I can go to any other big box CRM company and try to find something that I'm going to like. And I can tell you right now, I won't. I'm not going to find any of them that I like because there's not going to be a single one where I log in and this is my screen. Not a single one of them. All of them I'm going to log in and there's going to be some, you know, shit dashboard that, you know, tries to greet me and show me some rundown and