geekflow

dashboard

Inspiration

Collectors want to know the prices and market fluctuations of their games, cards, and figures. But constantly checking eBay or asking AI gets tiring fast.

What it does

With geekflow, you can manage your collections and get daily updates on prices and seller information. This way, you will always know value of your items, simplified with a stock-like chart system, so that you can optimize sales and purchases.

How we built it

Frontend: Python, Streamlit Backend: Python, FastAPI, Uvicorn Database: None (JSON based) APIs: OpenAI API, JSON Cloud/Deployment: None (local)

Challenges we ran into

We could not integrate the eBay API, since it required to wait a whole day for verification. We ended up generating data ourselves for the time being. Gemini API was too slow and often failed requests due to high traffic, so we switched to OpenAI API.

Accomplishments that we're proud of

We were able to create an usable and polished prototype app with accurate chart visuals. Managed to integrate OpenAI to create more accurate market readings, accounting for item quality and previous prices in its estimates.

What we learned

It was our first time creating a full-stack application using APIs and collaborating at the same time through Git. Was our first time creating an app with direct usage of generative AI.

What's next for geekflow

We first plan to fully integrate the eBay API to get real item and seller data in real time. In a full product with paid subscriptions, we would include the ability to create sub-collections, and have links to current sellers and AI summaries of price fluctuations, among other features to make selling and checking item descriptions more easy.