#199 - If You Think It. You Can Build It.

[00:00:00] Here's the question. Here's the question I have for you guys. How curious enough are you to try? That's what this comes down to. Not whether you know how to code, not whether you have all the answers, not whether you know AWS or GitHub or terminal commands. Those things can be learned. All that shit can be learned. The real question is, are you curious? Are you willing to play? Are you willing to experiment?

[00:00:27] Are you willing to be bad at something for a little while? Are you willing to ask better questions? Are you willing to imagine something that does not exist yet? Because I think only a few of you listening to this will actually do it. If there's 100 of you listening today, when I roll out this episode, five of you are gonna do something. Only five of you will actually do this. A lot of people will hear this and think, that sounds interesting.

[00:00:55] But only a small number of you listening will hear this and say, I'm in. I'm frickin' in, Andy. And those are the people I wanna find. Who are you? Those are the curious ones, the creative ones, the courageous ones, the ones who don't wanna wait. The ones who want to build. The ones who want to take the power back. Because that's how I see this. This is how we take back the power. This is how we stop waiting for the industry to change.

[00:01:25] And start changing it ourselves. This is the time. The time is now. If you are an irrigation professional, old or new, who designs, installs, or maintains high-end residential, commercial, or municipal properties. And you wanna use technology to improve your business, to get a leg up on your competition.

[00:01:50] Even if you're an old school irrigator from the days of hydraulic systems, this show is for you. Welcome back to the Spring Cleaner Show, my friends. This week, I wanna talk about something that I'm super fired up about. And I'm using the word super fired up because I don't mean kind of interested in this topic. I mean, I am frickin' fired up. I wake up thinking about it. I think about it all day. I go to sleep thinking about it. Every moment I have, I am practicing this new craft.

[00:02:19] Because I think this could be game changer. Not just for me. Not just for software developers. But for you. Irrigation contractors, distributors, manufacturers, service companies, entrepreneurs. Anybody in this industry who has ever had an idea and thought, Man, I wish this existed. Or, why doesn't somebody build this? Or, I wish this software worked differently. Or, I wish a manufacturer would add this feature.

[00:02:48] Well, here is the shift. Here's the opportunity. Here's where we are right now. You may not have to wait anymore. You may not have to wait for that good old fashioned traditional irrigation manufacturer to put your idea on some imaginary internal feature list. You may not have to wait six months or a year or three years for someone else to decide

[00:03:15] your problem is important enough to solve. You may not have to wait for permission. Because now, if you can think it, you can build it. I seriously mean that. If you can think it, you can build it. And that's why I think creativity and curiosity is the spark that lights these fires. If you can think it, you can build it. And I know that may sound like a big statement, maybe even kind of ridiculous.

[00:03:45] But I can honestly tell you that I'm four to six months into practicing the craft of vibe coding. And here's what I can tell you. Something has changed. Tools have changed. Our access to these tools have changed. The speed at which you can implement things has changed. And maybe most importantly, it's the power that has changed. Because now, you have the power. You are the power.

[00:04:13] You are now only limited to your curiosity and your imagination. And I want to challenge you. How curious are you? How curious can you be? What can you dream up? What problem have you been putting up with for years because you assumed there was no other option? Or you sat there and you waited for manufacturer A, B, or C to add some feature into their controller software,

[00:04:40] only to realize they're never going to add the damn feature. They don't even validate your need for this feature. What have you wished for that doesn't exist? What internal process just drives you nuts in the company that you own, the company that you work for? What customer experience could be better? Or what tool would make your team faster? What would you build if you stopped waiting for someone else?

[00:05:08] And that's what this episode is about. And because I'm speaking to you and I can't show you, this is not going to be a super technical how-to. We'll save that for later. I'm not going to walk you through every single terminal command that I use to Vibe code. And I'm not going to pretend that this is magic and you just press one button and there's no learning curve. It is definitely technical and there's definitely a certain foundational level of knowledge required.

[00:05:38] But I can also tell you, if you're willing to learn and you're curious enough to keep asking the questions on how something works, you can definitely do this. Okay, so there is a learning curve. But I want to give you a behind-the-scenes look into how I approach Vibe coding, how I think about it, how I start, and why I think this might be one of the biggest opportunities in front of curious people right now. This is not the biggest opportunity for all people.

[00:06:05] For a lot of people, this stuff is right over their heads. They don't get it. They don't understand it. They're afraid of it. They think it's going to take their job away. And you know what? For those people that think that, they're right. Because generally, you are right about whichever way you think. So you get to choose how you think. But if you are the curious one, you can participate. Okay? And when I say the curious one, again, I mean you. You are the contractor.

[00:06:33] You are the service manager, the distributor, the manufacturers rep, the entrepreneur. It doesn't matter what your particular trade is. Be curious in whatever it is you do during the day. It's the person who sees the problem when other people walk past it. Okay? The person who says, there has to be a better way. That's who this is for. So let's start. What is Vibe coding? And what do I mean by Vibe coding?

[00:07:02] So to me, Vibe coding is using AI as a creative and technical helper or partner. And in this case, to build a piece of software. You don't have to sit down and write every line of code by hand. You don't have to know everything. You don't have to have a degree in computer science. But what you need is simply, again, curiosity. You need to be willing to experiment. You need to be willing to describe what you want.

[00:07:29] You need to be willing to test things, break things, fix things, see how things work, and keep going. Vibe coding is not just telling AI to, you know, build you X, Y, and Z app. That's not really the point. The point is learning how to think through a problem, explain the problem, shape the idea, and then use AI tools to turn that idea into something real.

[00:07:52] Like literally a real piece of software on the real internet, on a real domain, with a real user login. All that shit. Seriously. So I'm about four months in, I think, sometime this winter that I first just started to dabble. And I have no expert, but I've built a handful of apps. And I'm going to walk you through one of them in particular here today on this episode.

[00:08:18] But the first one, one of the first ones I built was the slide, and you can see it at slide.sprinklernerd.com. That was one of those projects where I had an idea. I wanted to test it. I wanted to see if I could actually build something useful. But then another project came up. And when I say it came up, it's actually always been there. This is one of those instances where I was wanting to scratch my own itch.

[00:08:46] And the app concept is Branchboard. Branchboard, all one word. You can visit the website branchboard.app, and that's .app. And this app came directly from a problem that I have had since I started sprinklersupplystore.com. And this app was built for sprinklersupplystore.com specifically.

[00:09:10] And how we interact with our and communicate with the irrigation distribution branches that ship product for us. So when I started Sprinkler Supply Store, I realized something pretty quickly. And that was email is a terrible communication tool for operational communication. Especially when you're trying to communicate with remote branches. And it dawned on me that what email really is, is a one-to-one communication. It goes from one person to one person.

[00:09:40] But then what happens if that person is out of town? What happens if that person's on vacation? What if they're at lunch? What if they leave the company? What if requests get buried in their email box and they forget about it? What happens if no one else at the branch knows that the email you sent to the branch exists? That's the problem. Let's say we send an email to a branch and it says, cancel order number 63574. What if they don't get back to you? Did they get the message? Did they cancel it? How do you know?

[00:10:09] It's a one-way to one-person communication tool. There's no feedback. And that's what the problem is. I wanted to go from trying to message one person to messaging the branch so that all eyes could see the message. So that's really the difference. And that was the problem that I wanted to solve. So again, I've had this issue. That's not an issue. I've had this problem for 15 years.

[00:10:35] And back in 2011, when I couldn't find a specific solution for it, I did find a company called Smartsheet. And it's kind of like a Google Sheet, a little bit different feature set. But it's like a Google Sheet. And with the one added feature that we could have a conversation about a line on the sheet. Almost like having a communication thread inside one action item on the sheet. And so that worked. It worked for 15 years.

[00:11:05] We're actually still using it now because we're in transition to go from Smartsheet to Branchboard. But because I couldn't find a piece of software to do exactly what I needed, I was like, yeah, F it. If it's not out there and I've wanted it for this long, I'm building it for me. You know what? Maybe I'm not the only one out there. Even if I am, that's okay because I need the tool. I need this software. So here we are. Fast forward to this year, vibe coding.

[00:11:35] I'm like, Andy, you're going to build this tool. And so I asked myself, could I build the software I always wanted? Could I scratch my own itch? Could I build the thing that didn't exist? And then I thought, screw it. I'm doing it. So I did. Welcome to Branchboard.app. The idea is simple. Stop emailing people. Ask the branch. That's the whole concept. Instead of sending a message to one person and hoping they see it,

[00:12:02] Branchboard creates a request that belongs to the branch. The sending branch can see it. The receiving branch can see it. Someone can claim it. Someone can comment. Someone can upload a photo. Someone can resolve it. And now there's a trail. What was asked? Who answered? How long did it take? What was the outcome? That is the kind of simple, specific problem that vibe coding is perfect for.

[00:12:31] So we got to start with the problem. And here's how I approach it. I don't start with code. I start with the problem. And I would encourage you to do the same. Don't start by saying, I need an app. That's too specific. Maybe you don't need an app. Maybe you do. Start by asking, what is the pain point? What is inefficient? What keeps happening over and over? Where are people wasting time? Where is information getting lost?

[00:13:00] Where are customers frustrated? Where is your team doing manual work that should be easier? You don't need to solve it yet. Just go deep into the problem. Write it down. Describe it. Talk it out. Get clear. Because the better you understand your problem, the better an app will be at solving it. And this is where I think AI becomes really powerful. You open your favorite AI companion.

[00:13:27] Could be for me, I use ChatGPTs. We'll use that in this example. Instead of telling ChatGPT what to do, start differently. Start with a brain share. I like the word brain share better than brainstorm. You can use whatever word feels right to you. I don't mean using that word in your chat session. I just mean for the concept of imagination and creativity and being curious with your AI agent.

[00:13:56] I just like to describe it as brain sharing. And you can use whatever word feels right to you. The point is that I'm not just giving AI instructions. I'm using it as a tool to think with, to explore ideas with, to ask questions, and to challenge me. I'm letting it suggest possibilities. So, let me give you an example of what I might ask ChatGPT. Here we go.

[00:14:26] I have a problem I'd like to brain share with you. Together, I'd like to conceptualize a web application that could solve this problem. Maybe other people in the world have the same problem. So, maybe this could become an app for more than just me. Then, I explain the problem. So, in this case, for BranchBoard, I'm going to tell you exactly what I asked ChatGPT to help me with. Okay? Here we go.

[00:14:54] Email is a terrible tool because it relies on one specific person. It's one-to-one. I want to think about a better way for branch-to-branch communication. Imagine a company with hundreds of locations. Branch32 wants to ask Branch371 about a product transfer. Today, they might call or email someone. But what if that person is out of the office? No response.

[00:15:24] How might we build an app that behaves more like a task list, a question board, or request board, where everyone in the branch can see the request and anyone can take action? Okay? That was the thought theory concept I asked ChatGPT. So, that's where it starts. It didn't start with code. It starts with the problem and asking for help.

[00:15:50] Brain sharing different ways that we might solve that problem. Okay? And then you might say, how might we design a web application that solves this? And then you see what your AI agent comes up with. You keep asking your agent questions. You keep refining it. You keep brain sharing with it. Let it be curious. Let yourself be curious. See what...

[00:16:21] See what... Gosh. See what you can come up with. And eventually the narrative starts to become clear. You can kind of feel it. You'll know when you're reading through the possibilities of what this application that you're dreaming up can do. You'll know when it's ready. Okay? So, that's the first step is to brain share with AI. Talk through your problem. Brain share the possibilities.

[00:16:47] Get clear on what is the best way to solve that particular problem. Step two, which is... And I might get these steps out of order. So, this technically might not be step two. It's just the next bullet point on my podcast episode list here is the PRD. Okay? And those are three letters you're going to want to become familiar with. And you're going to want to know what it means. PRD. And that stands for Product Requirements Document.

[00:17:16] You need a PRD. And this is basically a blueprint for the app. Okay? A PRD tells the coding AI, like Codex or Claude Code, what the app is, what it should do, how it should work, what features should exist, what should not be built yet, what the user experience should feel like, what the technical stack should be, and how to phase the build. And this is where a lot of people go wrong.

[00:17:46] They jump straight into coding with some vague idea. But vague inputs creates vague outputs. If you want to build a good app, you need a good set of build instructions. And that's where the PRD comes in. The PRD becomes your instruction manual. So, for Branchboard, the PRD said things like this. And my PRD, it was freaking extremely long. It would probably take me 30 minutes just to read you the PRD. Okay? So, I'm not going to do that.

[00:18:16] But I want to give you some snippets into what goes into the PRD. Okay? So, here's a couple of things that were in the PRD. The app name is Branchboard. The tagline is, stop emailing people. Ask the branch. Branchboard replaces person-to-person email with location-owned operational requests. A user can create a request and send it to a branch. Everyone in the sending and receiving branch can see the request.

[00:18:44] Someone can claim it, comment, upload photos, change status, resolve it, close it, and generate a read-only public share link. That's the product summary. Then we define the first use case. Sprinklersupplystore.com is communicating with SiteOne branches that fulfill and ship orders. Then we need to define the design principle. Simple enough for a branch counter. Sexy enough for enterprise. I love that.

[00:19:12] Because it tells the AI agent what we're going for. Simple, premium, modern, not overbuilt, not ugly, not some generic admin tool. The PRD lists what we are not building yet. No Shopify integration yet. No SMS. No billing yet. No gamification. No advanced enterprise features yet. And that is important because your AI agent will sometimes try to overbuild something. So you have to tell it what not to do.

[00:19:40] So the MVP should stay focused. And again, MVP is minimum viable product. So for branch board, the MVP is branch-to-branch requests. That's it. The PRD then gets more technical. It starts to define the front end, the back end, the database, the hosting, the AWS resources, the login system, the API endpoints, the data model, the user roles, the request categories, the analytics, the UI pages, the phase build order.

[00:20:10] Now and again, this episode is not... And again, this episode is not a technical how-to, so I won't read the entire PRD. But here's the point. The PRD takes the idea out of your head and turns it into a detailed plan that an AI coding agent can actually build from. Okay? Your idea becomes a blueprint document. And the document becomes the instructions. And the instructions become the app.

[00:20:38] So as soon as you get through brain sharing with, call it ChatGPT, then you're going to ask ChatGPT to build you the PRD. And ChatGPT will know exactly what the PRD means. Okay? So that's what I challenge you to do next. Tonight, this weekend, just have a brain share conversation with ChatGPT and then ask it to produce a PRD and see what it comes up with. And then do a little get curious about what's in that PRD.

[00:21:05] And if you don't know exactly what it's talking about, go do your own research so you can learn about what this PRD is, what it means, and how you're going to use it. So now let's briefly talk about the tools. Again, this is not a technical tutorial because I can't share my screen, but I will share my setup. I use ChatGPT to think through the ideas and create the PRD. Then I use Codex from OpenAI inside Visual Studio. Some people use Claude from Anthropic.

[00:21:35] You also need to have a GitHub account and at least some familiarity with how GitHub works because GitHub is like the middle manager where it tracks all of your code changes and it sends them up to, let's say, AWS. And so it's extremely important to know how GitHub works. And for my apps, I've been building with AWS. Some people have been using Vercel. There's a few different choices out there. You might choose something else. And there are different ways to do it. So my setup looks like this.

[00:22:04] Visual Studio Code GitHub. And then Codex runs inside Terminal, which is inside Visual Studio. I use AWS for hosting and backend infrastructure. And this means you need to get a little bit comfortable with Terminal. Not 100% of the time, but a lot of the work happens inside Terminal, which is typing commands. And when your agent is off building the code, you're going to see the commands in real time.

[00:22:30] It's kind of fun to just sit back and watch Codex and Cloud Code do their thing because it does look like waving a magic wand. It's pretty amazing. But don't let that scare you. You don't have to know everything before you begin. You can learn just by doing it, being curious. And you can even ask your coding agent what something means, and it's going to tell you.

[00:22:57] And honestly, I think that sometimes the best way to do it and to learn is just get in there and get your hands dirty. And I don't know what word we can use for coding in computers because your hands won't be dirty. Maybe just get your hands, keyboard sweaty, sweaty, sweaty fingers. Okay, so you got to give the PRD to the coding agent. And so you copy the entire PRD.

[00:23:24] You paste it into, in my case, I paste it into Codex. And then I tell it to begin. And the PRD is detailed enough that the agent knows what to build. It will start creating the files, folders, components, routes, backend services, database structures, deployment scripts, and everything else it needs to build the app. And along the way, it's going to ask you questions back in Terminal. It may need you to approve certain commands. It may need to give you permission to install different packages.

[00:23:53] You may need to make decisions when something gets stuck. It's not totally hands-off. And I think that's important to say that it's not just complete automation. Vibe coding does not mean you disappear, go away for six hours, and come back, and it's magically done. It is kind of like having an assistant on your desk. You are still in charge. You are still the director. You are still the product owner. You are still the designer. You are still the person with the vision.

[00:24:22] And the AI is just helping you build. But you still have to guide it. And I think that's what makes it powerful. What makes it possible is that you don't have to be the person who knows every technical detail. But you do have to be the person who knows what good looks like to you. What are you going for? Is it meeting your needs? You have to know when it feels right. You have to know when something is too complicated.

[00:24:50] You have to know when the user experience is wrong and it's off. You got to keep steering the project. That's the vibe part. Okay, so you guys might be confused now, but that's okay. It is a little bit confusing. But what I want you to know is during this process, none of this, depending how you set it up, but typically none of this is going straight up to the internet on some domain live right away. It's being built local. It's being built on your machine locally.

[00:25:21] That means it's not on the internet. It's not live. It's not on the domain. It's running on your machine. And usually what you're going to do is it's going to give you a URL where it's listed locally and it's on a URL and a port. And you can just copy that, paste it into whatever browser you use, and you can view your app. Just like viewing it on the internet, but it's just viewing it local on your computer. And that's where you test it. You can look at it. You can click around. You can see what's broken.

[00:25:48] You can do all this stuff locally on your machine without deploying it up to the internet. Okay? And then once the app is running locally, this is where the fun really starts. Because now you can start to make changes and tweak things. And yeah, this is where it's almost like everything else that has come before this has been building the rocket ship.

[00:26:17] And once the app V1 is done and it's live on your machine, that's when it's time for liftoff. Because making changes, I think, is much where the magic happens. And again, this happens all in terminal. You can just say things like, change the left navigation from settings to admin. Then the agent goes to work. It finds the files. It makes the changes. It updates the app. You test it. And that's a small change.

[00:26:46] But you can also make bigger changes. You could say something like, I want the irrigation tech to be able to upload an image and assign it to a zone in the sports field. Now that's a feature. And the agent can help you think through the data model, use the interface, the upload process, the database, and how the feature should work.

[00:27:10] And that's where your imagination and your curiosity and your creativity start to matter. Because if you can describe it, you can begin to build it. Maybe not perfectly the first time. Maybe it's got some errors. Maybe it didn't quite work. But you can just keep tweaking it. And this is why I think it matters for this industry, for the irrigation industry, and perhaps all service businesses, and maybe just general in the world.

[00:27:39] But here's why I think it matters. Because that's why I'm talking about it here. This industry has a lot of problems that are specific. Very specific. And sometimes the problems are too niche for big traditional manufacturers to care about. Sometimes they don't understand the real field well enough.

[00:28:07] Sometimes they don't understand what happens between the technician, the office, the distributor, the customer, the controller, the pump, the sensor, the water meter, and the actual landscape. They don't understand the weird edge cases. They don't understand the phone calls. They don't understand the truck rolls. They don't understand the branch counter. They don't understand the service manager trying to figure out if a part has shipped. They don't understand the customer asking why their grass is brown.

[00:28:32] They don't understand the contractor trying to manage 3,000 irrigation accounts with tools that were never really designed for that exact job. But I know that because you are listening to this podcast, you understand it. You see it. You live it every day. That means you have the advantage. You have the domain knowledge in your field of expertise.

[00:28:59] Even if you are a sales rep for a manufacturer calling on distributor branches, you have the domain knowledge on what it takes to be the best manufacturer sales rep calling on distributor branches. What tool do you need that no one else is paying attention to because you have that domain knowledge? And domain knowledge plus AI coding tools is incredibly powerful because now the person who understands the problem can participate in building the solution. That's the shift.

[00:29:26] For all these years, you either had to wait for someone else to build it or you had to raise money, hire developers, create a software company, and hope you could afford to build the thing. But now you can prototype. You can test. You can build tools. You can build customer portals. You can build calculators, dashboards, scheduling helpers, sensor views, quoting tools. You can build whatever the F you want. Maybe some of these tools stay internal to your organization. Maybe some become products.

[00:29:54] Maybe some become training tools. Maybe some become part of how you differentiate your business. But that's the opportunity. So, the curiosity test. Here's the question. Here's the question I have for you guys. How curious enough are you to try? That's what this comes down to. Not whether you know how to code. Not whether you have all the answers.

[00:30:21] Now, whether you know AWS or GitHub or terminal commands. Those things can be learned. All that shit can be learned. The real question is, are you curious? Are you willing to play? Are you willing to experiment? Are you willing to be bad at something for a little while? Are you willing to ask better questions? Are you willing to imagine something that does not exist yet? Because I think only a few of you listening to this will actually do it.

[00:30:50] If there's 100 of you listening today, when I roll out this episode, five of you are going to do something. Only five of you will actually do this. A lot of people will hear this and think, that sounds interesting. But only a small number of you listening will hear this and say, I'm in. I'm freaking in, Andy. And those are the people I want to find. Who are you? Those are the curious ones. The creative ones. The courageous ones. The ones who don't want to wait.

[00:31:18] The ones who want to build. The ones who want to take the power back. Because that's how I see this. This is how we take back the power. This is how we stop waiting for the industry to change and start changing it ourselves. This is the time. The time is now. So here's my invitation. If this episode speaks to you, message me. Seriously, message me. I say this a lot. People do message me and I appreciate it.

[00:31:48] Here is my personal cell phone number. 208-908-3229. And even if I already know you and you already know me, message me anyway. And say, I'm in, Andy. Teach me vibe coding. That's it. If I can get, let's say, 20 of you, I will host an online vibe coding class where we will build real apps together. I'm serious. I'm not just putting this out there. In theory, this is not hype.

[00:32:14] We will build real apps to solve real problems with your real ideas. We'll imagine stuff. We'll build the PRD. We'll use AI. We'll build it. We'll test it. We'll learn. I'm serious. Please message me. 208-908-3229. Because I believe there is a curious army out there. People who are all in. People who are ready to change this industry from the inside out.

[00:32:41] This industry is not changing by the big name brand manufacturers anymore. This industry is going to change because you are going to change it. And I want to know who you are. So as I wrap up this episode, I want you to know this. Use your imagination. That's what vibe coding is all about. That's the fun part. Be curious. Try something. You will be surprised what is possible. If you think it, you can build it.

[00:33:08] And seriously, if you think it, you can build it. You no longer need to wait for someone else to decide your idea matters. You don't need to wait for a manufacturer to add it to their feature list. You don't need to wait six months for a release. You can build it now. You can build it fast. You can build it exactly the way you want it. You have the power because you are the power. You are now as strong as your curiosity and imagination.

[00:33:37] So test it for yourself. How curious are you? How curious can you be? What can you dream up? And more importantly, what are you going to build? This is The Sprinklyner Show. I'm Andy Humphrey. Stay curious and go build something, my friends. I can't wait to see what you build.