How To Build and Sell Mini Apps With AI ( Full Practical Guide )

Key Takeaways

Recently, I discovered something interesting. Instead of building online tools and micro SaaS and products and get stressed in marketing and competing with everyone online, you can build simple tiny applications targeting local businesses in your area in your country. And the good news, with the help of AI, you can do this in literally hours, not days or weeks. In this video, I will show you step-by-step practically how you can do that. And if you follow up, by the end of this video today, you can have your own product online. I'm sharing this because recently, last month, I built two applications for a local car oil company. The first one is simple app to track their employees and work, and the second one is a simple accounting application. And if you check it on mobile, it's fully responsive and looks like a mobile application. I'm so excited about this video, so if you are ready, let's get started. Okay, without wasting time, let's get into work. So, first go to Google and search for GitHub desktop. Click on the first link. In case you don't have GitHub desktop, just download it. It's a simple application to manage our projects. Now, in my case, I already have it, so I will open GitHub. And now, simply click on file, new repo, and give it a name, for example, local mini app, anything, and click on create. Now, of course, the first time you need to create an account. It's a free account on github.com. Just go here and sign up and connect with the application. I don't want now to go in these simple details. It's somehow easy and we have a lot of tutorials online about. So, simply get a GitHub account and create a new application. Second, we need to install Python in case you don't have it. If you are following me, you should have Python. We have built a lot of projects before on this channel. So, in case, just go to downloads here and download and install Python. After that application again, we need an IDE, a place where we build our applications. So, we need VS Code. Again, in case, just download it and install it. After you do this, we are ready to start building. We are going to open the project inside VS Code. So, now we have an empty project. The second step is you need to install an AI coding agent inside VS Code. For example, in my case, I use Claude Code, this one. You can go here to extensions and search for Claude and install Claude for VS Code. Or, you can use Clowne if you want. We talked about before. Or, you can use, for example, CodeX, ChatGPT. I have it also, but I prefer using Claude Code. Perfect. Please now focus very well. The number one mistake everyone makes when coding with AI, they just go here into AI and tell it build me an application about blah blah blah. Never do this. Before you jump into prompting, you need to envision and plan your application. So, you need to understand, maybe on paper, you need to draw on Canva, whatever you want, you just to understand the structure and how the application will work. Then, you go into AI and start prompting. In my case, to make things practical and show you how you can build something real right now, I will go with a simple booking application for local shops. Please focus now. This may be the most important part in this video. I will use this powerful prompt. I will start by saying I want to create a new Python Django web application. So, first you need to decide on the framework, which programming language and which framework you want to use. And I recently discovered that Python Django is one of the simplest ways to build web applications fast with AI. Then, we have the tech stack. I'm defining here what I will use in my application. Again, I will use Django templates with Tailwind CSS. So, I'm telling the AI I want to exactly use Django templates with Tailwind to style my website and my web application. Second, and this is one of the most important things when it comes to building with AI, you tell it to build in a modular structure for easy updates and maintainability. Then, we have the app should have one back end with two front end views. Since we need the application to be fully adaptive and responsive on mobile, I will create two views, one for desktops and one for mobiles, but with a central back end. So, we have what we call single source of truth. We have two UIs connected to one back end for functionalities. You see here, both share the same back end for a unified source of truth. Then, since Django has built-in administration and this is super powerful and saves a lot of time, I will tell it use the built-in Django admin for all our admin and management tasks. It saves a lot of time. I will use SQLite with the Django built-in ORM. So, again, Django has a built-in ORM or functionality to work with databases easily. And here, I will tell it to use my existing virtual environment. We'll create this right now in few seconds. Then, I will use Resend for sending emails. Since we are building a booking application, we are going to send emails to users or to admins. And one of the best and easiest services to use is called Resend. Let me show you this. You can start for $0 per month. It's totally free for 3,000 emails. It's way more than enough for you to build the application. And I will tell it to enable terminal emails. So, when you are developing, you can see emails directly in the terminal when it's sent. It's great for development and testing. Then, for deployment, I will use Docker and I will publish on Coolify. You will see how easy it is also when we finish the application now in this video, how to publish it online in just few minutes. And now, we jump with development workflow. We will build the application step-by-step starting with creating the project structure. So, I'm telling it how to build the application. It's very important when you build with AI to go step-by-step, tiny steps, test, ensure it's working, then jump in the second step, and so on. And I will tell it to create a project plan. In case you want to stop and continue later, we have a checklist of all the tasks in a file inside the project, so you can continue anytime and you have a full plan for the project. Then, we have the application description. I will give it a name, for example, I will say simple booker, anything. And then, we add the features. This is why I told you you need to envision and plan the application before. So, here I'm telling you what I have. We have a simple calendar grid, the public page, and we have the admin panel. So, I describe what exactly I want from this application. We have the email system. We send a confirmation email, and so on. So, here you explain what do you want, what features you want in the application. And at the end, before you start, ask me clarifying questions to fully understand what I want. This is very important. So, before the AI starts planning, it will ask you questions to understand exactly what you want, and this will save you a lot of time. Before you hit enter, make sure you are in the plan mode. So, switch to plan with shift tab when using Claude Code and enter. And you can see now all these. We have 13 questions AI is asking to make sure it builds exactly how you want. So, I will answer them fast. Okay, so I answered the questions and now let's continue planning. This step is really very important so that AI understand exactly what you want. While it's thinking, I will go here to terminal and create a virtual environment simply say Python -m venv venv. That's it. And you can see now it's testing if we have a virtual environment. We have the folder now. Say yes, and it exists. And now, it will plan the full application for you. Let's wait for a couple of seconds. Perfect. Our implementation plan is ready. Now, a small advice. I have been coding and programming and building application for 15 years. But with AI today, I'm learning a lot of new things. So, even if you are a developer, a programmer, or especially if you are new to building and coding, it's very important, really, to read and understand the AI plan, not the codes, at least the plan. So, read it, try to research about things you don't know. If you, for example, you don't know what is a CDN, just Google it or ask ChatGPT about it. If you don't know what is Docker, just do a simple research about it and enjoy learning while you are building. Do this and you will thank me later. So, now we have the full plan. I will simply say yes and auto accept, so AI can start building our application. You just now need to sit down, maybe do some doom scrolling. I'm kidding. Just wait for the AI to finish building your application. Now, while the AI is building our application, I just want to mention that I will leave this prompt in the description below. Just read it, understand it, and all you have to do is mainly change this part, the core features, and your application description. And anything, resource, link, website, anything mentioned in this video will be linked in the description below. And while we are waiting for the AI to build our application, I want to take like 1 2 minutes to talk about this business model, why it's somehow simple, why it would work for you, and how to market this type of business. And how I got my client last month. Now, if you are building a micro SaaS, a SaaS, an online tool, you may feel that the competition is very high because you're competing online with thousands of builders building somehow similar application. And you will need to go with content marketing, creating content, building in public, sharing, and so on. And many of you feel like stressed when it comes to getting your first users to app, public, micro SaaS, online tool. This is why if you think about it, and you target a local store, a local business in your area, in your country, there is almost no competition because you are reaching out, you are targeting specific store, specific place, and building application for that exact store or business. There's something here. Usually, 99% of local businesses know nothing about technology. They know nothing about building applications or AI or how these things work. So, if you learn how to build these applications, you'll have a great opportunity to help local businesses solve simple problems with simple applications, and you will be building them with AI, fully with AI in hours. And for the pricing, usually, it depends on your country. Like, in my case, I sell these applications for like $500, $600. Sometimes $1,000. It depends on the complexity of the application and the target business. But, you need to study in your area usually how much those applications are being sold considering your currency and the average price of such applications in your country. Now, to be honest, you know, I'm transparent. For my application that I built lately, the company reached out to me to build the application. But, in your case, if you want to build and market these applications, usually we use reach out. So, you need to send cold emails or reach out directly with mobile or maybe personally by visiting the store. And to discover businesses, usually we use Google Maps. So, for example, you go to Google Maps and you search for let's say barber shop in your area if you want to target barbers, and you will see all these shops in your area. You can click, and you will see at least their phone number, their website, their email, and you need to reach out to them providing your service. You may want to reach for 20, 30, 40, 50 to get your first client. That's normal, but this is how business works. Perfect project completed, and we have the full system ready. I hope so. So, now we can read the booking system, the email system, and read the summary of what the AI built for you. Again, don't forget, read these outputs and enjoy learning while building. So, let's activate the virtual environment, say V E N V, scripts, activate, and let's run the application. Simply say python manage.py run server to run the application and test it locally. Perfect, it ran, and let's simply click on this link, hit the control on your keyboard, and open this link, and perfect. This is our simpler booker application. You can select the date and enter your name and the phone number and the time you want, and simply click on book appointment. Wow, we have validation also, so let's do it again, and perfect. Booking request received. And if you go here again and go to developer mode, you will see it's fully responsive on mobile. Perfect. So, our main application is ready, fully functional. And now, since the application is functional, you can go back to your code and maybe add more features, change maybe styling if you want, and do whatever you want step by step. So, always when building with AI, make sure to finish the core functionality first, then you can change anything you want like styling or adding more features or adding more tools or whatever. And if you go to the terminal here, I I told you before we asked it in the prompt, if you remember, here, enable internal emails for development, you will see that when we submitted the booking, we got this email. So, this is the email is sent here inside the terminal, and which is good for debugging or development and testing. Great. I will not leave you here. I want to show you also how you can take this application and make it available public online so anyone can access from the web browser over the internet. So, I will go here to website called contabo.com. One of the best places, affordable places to buy a VPS server. You can see here we have a cloud VPS, and if you compare the pricing with any other company, you will see here we get four CPUs, we have 8 GB of RAM for around $5 per month. It's way cheaper than other companies like Digital Ocean, Linode, Vultr, and so on. So, I will click here on this VPS to configure it, and we have here two options. We have an update like 6 6 CPUs for $7. Okay, I'll go with this one anyway. Make sure you select Ubuntu and go down here, um simply enter a password for the server. Simply click on next, and if you are new to Contabo, just sign up to get and buy this VPS. I'm already existing customers, I will log in with my email. I run all my servers, all my services, all my websites, everything on Contabo, and this saves me, by the way, over $700 every single month. I explain this in detail in my self-hosting course. If you are interested, you can check it out in the description below. Let's now continue. So, it's loading, and this is the final order review. Simply click here order and pay and get the VPS. Now, in my case, I got one to save some time. You see here I'm running all the servers on Contabo. So, what you need simply after you get the server is to copy the IP address of the server. We are done from Contabo. Now, we need what we call an SSH client to connect to the server and install an application called Coolify to make things simple when deploying our applications. So, I will be using an application called Termius for connecting to my server. I will simply click here in new host and simply paste the IP address here and the credentials, which is root and your server root password, and simply click on connect. And we are in our server. Don't worry, we are just now going to run one command to install Coolify and then configure our hosting and our application. If you open here your web browser and go to Coolify install, you will see it is just simple one command here. I will copy it. Go here, paste. Right click to paste and enter, and the full setup is automated. Just wait a couple of seconds, maybe 1 minute, for the installation to complete. It's that simple. Congratulations, we are done, and we get this IP address. So, I will copy it and again open Chrome and open this IP in my browser, and here we are inside Coolify. And it will ask me to register, so um my email address and the password. Perfect, create account. We are done. Skip setup, and we are inside Coolify. Second step is I want to point a domain or subdomain to be used with my application. So, I will go here and select any of my sample websites or sample domains, and point it to my server IP. So, I will go here to the DNS section, records, add a record, and simply say here at, and simply say here test, for example, as a subdomain, and then point to my server IP, save. It's that simple. So, now we pointed our subdomain here, and we can start the installation. If you go back to our project, you will see that we asked it in the prompt, if you remember also in the deployment, to create a docker file to deploy on Coolify. So, already we have this docker file. All you have to do is to build it. So, I will go here and and click on build so we can build the application. But, to do so, you need to have Docker installed locally on your PC. So, open here Docker Desktop. If you don't have it, just download it and install it. It's super simple and free. So, wait for Docker engine to run. So, here is Docker. It's open. I close it now and just say here enter to build this Docker image. Wait for a couple of seconds. Perfect. It's done. Now, if you go here to Docker extension inside VS code, you will see here we have the local mini app deployed or built in the last minute. So, I will go here and click on push to Docker Hub and now it will be pushed to Docker with this URL here, you see? If you don't know what is Docker Hub, it's like GitHub but for Docker images. If you go here to Docker Hub, here you will see all Docker containers, public containers for thousands of applications online. For example, if you want to install Twin AI model, you can use this Docker image. We have here MySQL, we have we have a Grafana, we have a lot of Docker images which is public and free or you can install with Docker. Anyway, so the application now or the Docker image is pushed. Let's go back here to Coolify and click on projects and click on add project. Let's call it for example local app test. Continue and click on add a resource and here I will select Docker image and simply paste the Docker image name and save. Now, automatically it will load the details. Just go down here and select port 8000, which is the default port for the Django applications. And here, let's enter the domain name https test. wb4s.com, whatever now a sample domain. And then go here to environmental variables and click on developer view. If you want to add environmental variables, you see here in our code, in this section we have the .env and this is the Django configuration. I will copy it as it is and paste here and just make sure here to add the domain for allowed hosts, which is test. And application site URL is again here https and the domain or subdomain and simply go down and save. We are done. That's it. We configured our application on Coolify with Docker. Now, click on deploy. Perfect. Successful. Now, go here to links, open our subdomain and perfect. Our application is up and running publicly and you can see it's secured with free SSL, which is automated with Coolify. Again, I know maybe there's a lot of pieces missing in your head, but please at least go and try this. Build your first application, try to host it, learn this approach. It will be a great opportunity for you now and in 2026. I know like 99% will just skip this and say, "Wow, a nice tutorial." and just go and continue doom scrolling. Be the 1% and test this. Try this and learn this. If you have any questions, I'll be waiting for you in the comment section. And if you are interested to be a professional solo builder online, you can join the waiting list for my new program solobuilderschool.com. The link in the description below. Thank you and see you next week.