How I Coded ANOTHER Profitable App SOLO (step by step / from scratch / with AI)

Another day, another startup. Recently, I moved to Malaysia and from there, I've pretty much been spending most of my time coding my latest project. And it's been super fun hacking away at something new again.

And luckily, I've been documenting the whole journey from coming up with the app idea to building the app to getting that first paying customer. In this video, I'll show you every step of the process so you get an idea of what it takes to build and ship a profitable app from scratch completely solo. And feel free to stick around until the end to see how much money my app actually made.

My app idea is just a simple tool that allows me to download video transcripts in bulk from YouTube channels or playlists. And you're probably asking, why do I want a tool like this? Well, it's no secret that YouTube is a gold mine of knowledge and information, but it's all trapped in video format.

So, I'm building this tool for one main reason. I want a quick and easy way to extract all this gold from YouTube videos in text format so that I can feed this data directly into my own custom AI models to help me with everyday tasks like writing better YouTube scripts or creating custom AI agents to help me with my startups. Then after giving it some more thought, I know I'm not the only one who needs this tool.

Pretty much anyone wanting to train their own custom AI models with YouTube content is facing the same problem. Manual transcript extraction from YouTube is painfully slow and tedious and currently there aren't many good solutions out there. So then I thought why not try building this tool myself.

It solves a real problem I am facing and it could potentially be another stream of income if I execute and market it well to the other users who are facing the same problem. But before I start building, I need to do some quick market research to see what my competitors are offering. First, there's YouTube transcript.

io IO, which offers bulk downloading but requires a subscription with a credit system. So, I can definitely see an opportunity for my app to compete by offering a simpler pricing model. And from looking at their landing page, they report having over 350,000 users worldwide, which is a clear sign there's already real demand for this type of product.

This gives me even more confidence in my idea and extra motivation to compete with them to take a slice of the market. There are also other sites like Note GBT and Tactic, but those sites only let you download transcripts. one video at a time.

So, my approach will be simple. Create a tool with a simpler pricing model and a super clean interface that makes downloading YouTube transcripts in bulk cheap, quick, and painless. After scouting out my competitors, I was ready to jump into my code editor and start building.

But before I could write a single line of code, I had to plan out what the first version of my app would be like. Otherwise, I'll either end up building something that's not useful, or I'll constantly get distracted adding shiny new features that don't actually solve the root problem. I prefer using the SLC framework to plan the early versions of my app.

I'm not a fan of shipping MVPs that are buggy at launch and don't do the key one to two features they promise just for the sake of speed or shipping fast. The SLC framework stands for simple, lovable, and complete. And I think it's a much better approach than just building something that's minimum and viable.

For my app to be simple, it needs to have one straightforward path to solving the root problem, and that's extracting YouTube transcripts in bulk. This means users should be able to complete the task in as few steps or clicks as possible. Nobody likes using a halfbaked product for version one.

This means users should be able to download thousands of transcripts quickly. My app won't be feature- packed at launch, but it will be complete. This means users should be able to sign up, make a purchase, and immediately download transcripts.

This is the leanest possible version of my app and the goal is to provide value from day one and then iterate the product based on real feedback from users. Like I've said before, I'm aiming to ship a version one of something simple instead of a version 0. 1 of something complex.

Now that I have a basic plan in place, it's time to commit and secure a domain for the app. I decided to go with YouTube transcripts. io as it's short, descriptive, SEOfriendly, and doesn't break the bank.

Now that the domain was purchased, it was time to start building. For this project, I'm using the standard web text stack, React and Next. js for my front end and API, Superbase for my database and authentication, and Stripe to handle payments.

Normally, setting up my projects would take me a few hours to get everything connected and working. But this time, I am using Tempo to help me with the setup process, and I was able to get my OR, database, and payments all working in under 10 minutes. All I had to do was click through their guided setup, paste in a few API keys and within minutes I had a fully functioning project with all the annoying stuff correctly configured.

Then I gave a simple prompt to Tempo describing what my app does. And then it just starts generating the project files, documenting project requirements, and building out the UI. And honestly, the first version of the landing page tempo generated was pretty impressive.

It wasn't perfect, but it was a solid starting point with a clean layout and some decent copy. From there, I gave the AI a few more prompts to refine some sections, like updating the feature section to better highlight what the app actually does, adding a use cases section and an FAQ section to answer common questions or objections up front. Before we move on, let me give you a quick tour of the Tempo UI.

The interface is split into three tabs: product, design, and code. In the product tab, you can write out your product requirements, features, and even create full user flow diagrams of your app. The most useful thing about this is that all this information gets fed directly to the AI as context, which reduces hallucinations massively.

In the design tab, you can build out your UI just by clicking and dragging elements pretty much just like any other design tool like Figma or Web Flow, but it generates actual React code behind the scenes. They also support importing your custom Figma designs, so you can use them as a starting point and ask the AI to generate React code for you based on the designs. Finally, the code tab is where you can open and edit any file in your project just like you would in VS Code or cursor.

It also integrates nicely with GitHub, so you can push your code directly to a repo and keep everything version controlled. With the initial setup all done, I push the code to GitHub and fired up Cursor so that I can start building out the core features of my app. So, enjoy this time-lapse of me hacking away over the next couple days whilst I get this app ready for launch.

[Music] [Music] Oh, hey. [Music] Heat. Heat.

[Music] [Music] After days of non-stop coding and testing, I think the app is finally ready. It's time to launch and get some users. My plan for the initial launch is pretty basic.

I'll post on all of the typical launch platforms like Product Hunt, Hacker News, and Twitter to try my luck going viral, but I'm not banking on this too much. My tool is quite niche and solves a very specific problem. So there's really no point trying to appeal to a wider audience at this stage.

And hot take, I don't think Product Hunt is actually a good platform to launch for most indie hackers. Unless you're making a product specifically targeted to other founders because the truth is regular customers aren't going to Product Hunt to find tools to solve their problems. It's usually just a bunch of other founders looking for a place to share their products.

But you do get a nice backlink for SEO and a bump in traffic if you can rank highly. So, it's at least worth trying. That's why I'll be spending most of my time going directly to my ideal customers and building relationships with them first.

This means finding niche subreddits and forums and sharing my launch there instead. So, how did the launch go and how much money has my app actually made? Well, I can tell you my app didn't blow up overnight.

And I'm not going to lie, it was discouraging at first seeing my launch get less traction than I hoped. It's disappointing, but this won't be one of those videos where I launch my product and it magically goes viral. And honestly, I think sharing realistic launches like these can be more valuable than the overnight success stories because whilst they can be inspiring, they can also create unrealistic expectations and discourage you to keep going if you don't see immediate results.

So, even though the launch wasn't what I hoped for, I'm not going to let it distract me from improving and growing my product until I can get it to a point where I'm satisfied and feel like there's nothing more I can do. Since the launch, I've managed to get 13 user signups with six of them converting to actual paying customers. And so far, my app has earned over £200 since the launch.

It's not some life-changing money, but at least it's proof that people are willing to pay for my product, even at its earliest stage. So, what's next? I have a bunch of ideas and improvements I want to make to this app already.

And of course, I'll be documenting everything on this channel.