Hello everyone,
I’ll be honest with you — I don’t even know where to start, the project has evolved so much since the last post, and yet it hasn’t even been that long.
I’ll try to keep this update not too technical (though some terms will inevitably slip in), so that as many of you as possible can understand everything in the article.
Business Logic
In the last article, I told you that the market scanner was functional.
Well, now the automatic buying, automatic selling, user configuration management, account creation, deal assignment management, and a bunch of smaller tweaks are all fully operational.
So, it’s safe to say that the entire core of the application is now functional.
Here are a few screenshots from the console to give you a taste of what the app can do.
Here’s what the application logs look like at startup.
Goodbye JSON, Hello Real Database
I don’t remember if I mentioned it last time, but until now, all the data was stored locally in JSON files.
That was fine for the early stages of development — but absolutely not viable for making the app accessible to everyone.
It was a big step, and I honestly had my doubts, but the migration went super smoothly.
Of course, I ran into some issues and had to rethink a few poorly designed tables, but I’m happy with the result.
I decided to go with a PostgreSQL database hosted on Neon.
They offer a free plan for development and a pay-as-you-go plan starting at $5 per month, which is perfect for getting the app off the ground.
Now, the entire market is being tracked — price drops, price increases, and even changes in the number of cards available.
Server and API Setup
Today’s work was dedicated to setting up the server and the API routes that will power the user interface.
For speed of development and ease of use, I chose Fastify.
It’s clearly not the most powerful or feature-rich framework out there, but when you’re working solo on a project like this, sometimes simplicity is the key to shipping.
In this example, three deals were picked up and purchased at the best price, then relisted for a higher price within a fraction of a second — all without any user action.
Let’s Talk About the Interface
The user interface is the next big step.
I’ll be using Next.js and Tailwind CSS to build a clean and fast UI.
I also plan to rely on ShadCN components to avoid reinventing the wheel.
This library may slightly limit customization, but it saves an incredible amount of time and ensures high-quality components.
I’m aiming for a minimalist and clear design to make navigation as intuitive as possible.
For now, nothing is actually built on the frontend — I’ve only done the setup.
Here’s the list of pages I plan to include:
- Account creation page
- User configuration management page
- Main page displaying all deals and market movements in real time
- User dashboard
- Admin page
And what about you — would you add something else?
All the deals you see on this page were bought and listed 100% autonomously!
A Partnership Already Signed
I’m pleased to announce that a partnership has been signed with the very talented @deadzy . SPL Autobalancer — now the exclusive partner of the Splintershare marketplace!
By using this platform, you automatically recover 40% of the standard marketplace fees, ensuring even greater profits when using SPL AutoTrader!
Conclusion
There’s still a lot of work ahead — I’m under no illusions.
I’m still in the first phase of the 80/20 rule.
The polishing phase will inevitably be the longest, especially once I add logging, tracking, and gather feedback from beta testers, which will be crucial for the final quality of the app.
As always, stay tuned — if I keep progressing at this pace, it’s very likely that in the next post, I’ll start recruiting testers.