Update: after 145 hours and 3 months of work I feel burned out, if you have any opinions or ideas on where this project could be developed further in the future or for more serious things, feel free to write to me on Slack in partnership @Abro or …
Update: after 145 hours and 3 months of work I feel burned out, if you have any opinions or ideas on where this project could be developed further in the future or for more serious things, feel free to write to me on Slack in partnership @Abro or write to the site’s Gmail, thanks!
This project is all about creating themes and notes for shopping and expenses, which can then be analyzed. You can track exactly how much you’ve spent monthly, in the last 7 days, or in total. If that’s not enough, there’s also a built-in AI that can analyze your uploaded photos to offer tips and even compare two different shopping trips—for instance, to see how much inflation has affected your spending between the two.
I used Kimi K2.5 at the beginning, mostly for bug fixes. The first 30 hours of the project were purely solo-written, featuring solid user auth and backend logic. After that, I started using it again for the analytics—specifically the frontend and the data visualization—since I’d never done that before. Later, I switched to MiniMax M2.7 for the landing page, CSS animations, and the AI setup. I used it for the AI part because it was new to me, and for the landing page/CSS because I felt doing those from scratch was unnecessary and a waste of time. That said, I still wrote some of the CSS and landing page code myself, just not the whole thing. So, the AI models I chose—Kimi and MiniMax—aren’t exactly the top-tier, strongest ones out there, which meant I had plenty of well-deserved debugging to do after using them! :) . I used AI to make all the the README.md file
