Inspiration

I have seen many transactions tracker that claims to bring ease for tracking transactions, but tapping here and there just to create transactions stressed out users and will discourage them of continuing tracking their transactions with the app, this is why I thought carefully and realized this can be solved by integrating AI models like Gemini, to bring ease and convenience into expense/transaction tracking.

What it does

Trackr is an AI powered transaction/expense tracker, what the phrase "AI powered" mean to me is the integration of AI models to apps in such a way that users won't have to navigate through difficult steps. So trackr excel in this, just describe what you spent or received and it'll create it for you. You can also get insight on a particular transaction. Some of what trackr excels at is creating cards, transactions, giving insights and many more.

How we built it

I built this with flutter, for performance optimization, utilized riverpod for good state management, and nextjs for backend, I initially build the manual process without Gemini 3 integration so that I can get the clear idea of how it'll work, when I see that it is now working manually, so I added Gemini 3 which is a game changer for it.

Challenges we ran into

API key limit, Power issues, low end PC.

Accomplishments that we're proud of

  • Make use of riverpod for the first time, it is far interesting than provider.
  • Able to solve the challenges many people faced when managing their transactions/expenses.
  • Able to make use of one of the most advanced AI model(Gemini 3).

What we learned

I learned many things, but here are the noted ones:

  • Learned more about state management in flutter.
  • Learned more about flutter framework itself.
  • Learned more about AI models (Gemini 3)

What's next for Trackr

Pushing it to global recognition, so many people can benefits from what it offers

Built With

Share this project:

Updates