Build an AI App That Makes Money
Other Videos📅 2025-06-11
Transcript
[00:00] How do we build out an AI app that actually makes us money? In this video, I want to go over the tech stack associated with it, the tools, the type of artificial intelligence you should use, and from my experience, how long this will probably take you. That sound good? Let's jump in. So, let me show you how to build out these apps that actually make you real money based off my experience of building out the software bumpups.com. So, we're going to have three major things we're going to discuss in today's video. One, we're going to discuss how the heck we even do this, like what are the tools associated with this, which ones are optimal, and proceed. Two, we're going to go over
[00:30] strategy. What's the best way we can monetize our application, advertise our application, and everything we need to know when it comes to pitfalls that typical software developers fall into when building out an AI app. Also, to be clear, when I say AI app, this is like software or web apps. I'm just using it as an ubiquitous term, okay? This encompasses it all. And number three, let's go over the goal that you have for this AI app. Are you looking for an app that makes you maybe $2,000, $3,000 a month just so you can go anywhere in the world, travel, and be fine? or are you looking to build the next Google, the
[01:00] next big tech company? Because the way you approach building out the software is going to be entirely different depending on what you're trying to achieve. It's almost like, do you want to create the software that's like a mom and pop shop that has really, really good ramen, or are you trying to create the next McDonald's? McDonald's, please bring back the spicy chicken nuggets. I love those. I don't know why you keep taking them away. I'm pretty sure most people like those as well. If you don't know what I'm talking about, spicy chicken nuggets, spicy buffalo sauce. We're having a good time. Okay, step one here is tools. But don't worry if you
[01:30] decide to build this out in a noode way or use a noode app builder, strategy and goal will still apply in this logic. But let me go over tools assuming you want to take the leap here. Let artificial intelligence do the heavy lifting of coding and kind of just build out your application this way. So the first three major tools I want to go over here to build out an application is going to be your IDE, integrated development environment, cursor AI, VS Code, Windsurf, everything of this nature. your actual tech stack. And tech stack is a fancy way of me saying how are you creating the cake? Okay. Is your cake
[02:02] going to have a backend of AWS? Is your cake going to have a nice little frosting of React, which is a framework for the front end. Front end is what the user sees. Backend is what happens and where the magic happens where you provide value. And then finally, which is very important is going to be the AI. Right? If we're going under the preference that we're going to build out an AI app, leaning artificial intelligence to code out most of it and might be a surprise to you, but honestly at this point probably isn't a surprise to you. You can actually use AI to cut out most of it, even if you don't really have that much knowledge of coding. We are very much getting to a new age here where you can use regular dictation,
[02:34] just regular language and get out real code that actually works. Now, this isn't as simple as a hot pocket. You put it in the microwave for 2 minutes and then you have a nice little hot pocket you can eat. There's some work involved here, some learning curve. Once you get over that, pretty cool stuff can happen. So, IDE, and I'll leave a video in the description down below that has me diving in really in depth of all the different idees in the market and basically showing you which one to choose for what, whether you want to do no coding at all or you want to code a little bit more. But when it comes to actually where you going to build out this environment here, we need two major things. We need the ability to actually
[03:06] write out the code locally, eg just on your computer, like as if you made a Word doc that's not connected to the cloud. Locally, and cloud. So locally, which one do we want to use? Now there is a bunch on the market, right? We got cursor AI, we got Windsurf, we got VS Code. I love VS Code. This comes down to preference. Some people prefer using things like cursor AI and Windsurf as they want the actual AI to be in the IDE environment where they can ask questions directly. Now that video I referenced, I tell you why I don't like that. I
[03:36] personally prefer using a integrated development environment like VS Code. Not to say you can't do this with the other ones or a noode platform. Personally, I like coding in VS Code. I like keeping the artificial intelligence and the coding itself separate for most cases. Everyone's built different though. Like, do you like snowboarding or do you like skiing? Well, I really love skiing. Okay, you like skiing, you like skiing. I like snowboarding even though I find myself falling on the ground more, but I do like snowboarding more. So, VS Code is where we're going to be able to actually write the real code. And then over here in AI, I'll go
[04:09] over which models I personally like to use when writing code. Now, where's our cloud? What does a cloud mean? We're writing our code on our machine almost like we're writing in a word doc that's not connected to, you know, our Microsoft 365 system. We need something where we can put our code in the cloud so we can access our code anywhere on our laptop, other laptops, other desktops. In addition, this gives us version control. So, for example, let's say we have like version one of our app. We do something really crazy. We go down a little rabbit hole. You already know how these rabbit holes are for AI. They just sometimes they go down rabbit
[04:39] holes. It breaks our application. We need the ability to revert back to a older version very easily. And that's when the good old GitHub comes into play. I'll link that in the description down below that shows you a real quick 10-minute video of how to understand GitHub. But GitHub is fundamental. It is free to use. So is VS Code. Now comes to our tech stack. Your tech stack is going to be all the applications required for you to build out a product that provides value that the end user will actually pay for actually pay for it. Okay? Unless, as you see with your goal here, your goal might be different. So, let me
[05:10] give you an example text stack here. Now, I know you've seen sites like this where it's like go to a site, download a YouTube video, download a Tik Tok video, or if you haven't, there's these sites or like download the transcript of a video where it's very simple. The idea is that you simply put a link here, download it. You might be like, Corbin, why isn't this in English? It's cuz I'm in Indonesia right now. But the idea is this is an example of a software that typically would have ads on the side and you make money that way. So, to build out a software like that, but let's say the software's use case, which I might actually build out on this channel live in a playlist showing you step-by-step
[05:41] how-to. So, make sure to subscribe here for the purpose of getting timestamps for a YouTube video. Provide the YouTube link, get the timestamps out. The text stack would be this. We would use React for the front end. We would use Firebase and GCP for the backend to handle the actual logic of taking in that YouTube link and providing the timestamps after it. And then instead of us building out a custom pipeline, we could use bump ups here to get the API for time stamps. Eg just coming over the documentation here. Simply calling this URL in our application, user provides YouTube link, we get our nice little timestamps. Next
[06:11] part of the tech stack would be how do we monetize this? Now, if you plan on building a plan-based system, eg user comes and they have the ability to subscribe for $20 a month, $40 a month for a plan, this is where Stripe comes into play. And if you don't know what Stripe is, you probably do. It feels like they have a monopoly. like they just quite literally acquired Lemon Squeezy a couple months ago, maybe a couple years ago. Basically, Stripe, anyone that's trying to make money with software and AI apps, you kind of have to use Stripe. So, Stripe is going to be your way to monetize in a scalable way, a secure way. And to be fair with you,
[06:44] the checkout's very good, very good conversion there. Let's jump back over to the tech stack. In this context though, when building out this application, which I'll show in later videos on this channel, we use something like Google Ads. If you're familiar with this, you go to a blog and you see like those little ad spaces on the left column and the right column. We use something like this to monetize. Plus, we could maybe bid off some infra where the user puts in the link, watch this ad 5 seconds, click, get timestamps. So, React front end, Firebase and GCP for the back end, Bumpups API, Google Ads. That would build out a full-blown software application that we can start
[07:14] making money on. So, first step here for you to understand to build an AI app that you can make money with is build out a tech stack that's reliable, cost effective, and you like it. I personally, I know people like AWS. I'm not I'm not hitting on AWS. I just like Google better with the UI. I just like it better. Okay. So, for me, I typically lean towards GCP, Google Cloud. So, now comes into a very important question here, which is how the heck are we going to build it with AI? Here are three major models that I personally like when coding of AI. And I've done an entire series on the back end, an entire series
[07:46] on the front end showing you how I code of AI live in this kind of manner. You can check out those series. I'll make sure I either comment them or description them. Description them. I don't know if that's a word. And you'll be able to watch them, get full-blown context. All free. Don't worry about it. So, by the time you watch this video, these models may have changed names. As you know, with the AI space, it's not like the iPhone where it's like every year, it's like every couple months where it's like new model, new model. This is the best one yet. I like this. We got 04 mini. By the time you watch this, might be 0 06. 04 mini high and
[08:17] claude sonnet. Which one do I choose and which one I choose for what? I think you choose 04 mini for most common task as well with claude sonnet. These are very interchangeable. I personally like using 04 mini high when I need to do a more complex problem of code. But what you'll notice especially when you are doing the front end at least there is just a bunch of code where it's not even complex. It's just quite literally you just need the lines of code. So using a 04 mini model, you get it out fast. Same with clot sonnet. So opt for those. So cost associated with all this. What's really nice is that VS Code's free, GitHub's
[08:49] free. Obviously, if you use it more or higher usage, you could get into some paid plans. React is free. Firebase GCP is pay as you go. But what you'll notice is that the rates are pretty pretty generous. Bus API cost per use. Google ads, you're making money. Of Mini, O for mini high, quad sonnet, these are associated with those platforms. So whether you're paying for a $20 subscription a month or what they seem to be doing now, which is like they're going crazy. I don't know if you saw, but Google literally has a plan now for like 250 USD a month. That hurt my heart. But you could offer the $20 plan
[09:21] for both of these and you'll be fine. So now that we know the tech stack and how to build the AI app, and I've shown a ton of videos here pragmatically how to do that, what is the strategy here? Let me just give you some pointers here that is going to save you a ton of time that I lost with bump ups here, lost some months just because of the fact of that's the nature of building out a company. First pointer here is that less drawing on the whiteboard of what's going to happen, more MVP. This is a very common phrase in software. Ship it, ship it, ship it. Idea being you just got to get going here. Get the ball rolling. You are going to learn more about what's really next and what's
[09:52] really going to be the best possible avenue. The more you do, the more you build, because in reality, especially if you're starting at ground zero, you're going to have very limited context of who the true consumer is. This all just comes with data. And you know, data is king. So learning more about your consumer comes through consumers actually using your product. So MVP, get going. Next, which might seem like not that big of a deal, is name. The name of your product is a pretty big deal. Now, obviously coming down to your goal here, it might not be that big of a deal. For example, if I'm creating this tech stack
[10:23] here to do a web app that's entire purpose is just to give timestamps on a video or a YouTube video, then something like Tube Stamp is good. That name works perfectly because it hits the niche. So if the naming of your product and then kind of tailoring over to three here is kind of a very one-size fit all boom boom boom boom then your name can be very specific like tomb. But if you're thinking you know what I really want to iterate on this I want to do multiple years at this and I just want to build build and go have a name that you really love something that will last with you two years three years from now and the
[10:54] only reason I'm harping on this so much is because when we started this company it was originally called Tube Stamp and then we had to pivot to bump ups. There's a ton of stuff that's correlated with switching a domain name in software. It's not as simple as just, oh, just redirect it. Ton of stuff. And then the last part is go in a niche that you know to an extent. You don't have to know it perfectly. Lean into your strength. If you've been a graphic designer for the past 5 years, build an AI app that leans towards graphic design. You can leverage your network there. You can leverage different communities there because you've been
[11:25] there. Me personally, when creating bumpups and tube stamp, I leaned towards me creating YouTube content. I was like, you know what? I'm already in the space. I seen comments saying, "I want timestamps for this video. I saw no product solution at the time, mind you, 1.8 years ago. Create a tube stamp." Lean toward your strength. That's should be your strategy here. As if you go into absolute murky waters in an industry you have no clue about, it works. It's possible. But if it's just you or a small team or it's bootstrapped, you're basically trying to hop a 400 meter leg
[11:57] race on one foot or something. I don't know. Just you're slowing down. And then last one here is goal. What is your goal? I wanted to leave the last one here of goal because there is multiple ways you can approach creating AI apps and software. First way is Corbin. I'm just trying to have fun. This is purely a hobby thing and I could care less if I make money on it. It's purely just for me to build something, have something. It's fun. It's almost like honestly I feel like coding in general is like drawing and like painting but it's just kind of like a side thing and no hate on that. I love I love coding too. I think creating cool stuff that's completely
[12:28] unique just through text is a cool concept. So that's the first way. Let's just have fun. And most of the stuff I described here you can basically do for free if you just want to have fun. The second way is like that micro niche software where it's like we do one thing and we do it really good. Tube stamp. We provide timestamps for YouTube videos. We do one thing but we do it really good. That's our use case. Monetize it. The upper threshold of how much money you can make there is a little bit more limited because of how niche you are in the market. Which brings you to like the third or fourth way, right? The third or
[12:59] fourth way is are you approaching this with the idea of I want to build a multi-billion dollar company or a multi-million dollar company or do I just want to build a company that makes me 2,000 USD a month? The way you approach all three of those is going to be a different way. If you're trying to build the really big companies, you're going to need a team. You're going to need multiple engineers either through equity, through uh, you know, network, whatever it may be. In the beginning, it could be you and one other person, but long term, you're going to have to build out a whole team. Realistically, a lot
[13:30] of times people can find a sweet spot between you, one other person, or maybe two other people, make good money, and be good to go. That's up to your discretion. Raising money, that has its implications. How do you want to approach this? Personally, there's levels to it. I think you approach it with the mentality of, you know what, Corbin, I just want to quit my job and I only want to make like $2,000, $3,000 a month. You could live in a bunch of cool places in the world on that kind of budget. Like right now, I'm in Indonesia. Indonesia, if I can speak. And the amount I'm paying here comparative to what I paid in Texas is
[14:03] just way less. So, keep that in mind on your goal for building out this. So, that covers how to monetize and create money with an AI app. very very overarching way of doing it. This is quite a journey. This is quite a way, but I think it's pretty beneficial regardless of what path you want. So, make sure to check out other stuff on this channel here. More content to come. As you already know, I'll see you in the next