Hot Cocoa ☕️

Inspiration

Data is the currency of the digital era. With more and more amounts of data being generated and recorded, one crucial aspect is missing from most small business and startups - the ability to use that data to produce insightful results. We identified this issue and worked towards building an application which allows small businesses or even individuals to harness the power of data and forecast profits from existing features. This birthed: Hot Cocoa ☕️.

With Hot Cocoa ☕️, we give small businesses the ability to predict their future profits before they even sell a product or service. This has immense benefits for businesses and companies who do not have the technical skills to analyse their data.

What it does

The user is prompted to upload their business' data in csv format, which is used to train a machine learning off the whim. Then the machine learning model is used to predict or forecast profits for the business using RandomForest algorithm. In order to predict profits, the user needs to enter data regarding which we will forecast the profit. After data entry, the user gets predicted profit with a click of a button.

How we built it

We used Flask and React frameworks and used Python/ HTML/CSS for creating the project.

Challenges we ran into

Since we didn't have much experience in AI and ML, it took us a fair amount of time getting used to with sklearn and understanding RandomForest (the algorithm we used). Plus, we had to learn a ton about flask, the web framework we used for building out the backend.

Accomplishments that we're proud of

We are proud of building the entire frontend and backend along with the machine learning part of the application in under 36 hours, especially considering the fact that most of us were inexperienced with backend programming and machine learning it was a pretty good learning experience.

What we learned

We had many learning outcomes from workshops and self learning about AI.

What's next for Hot Cocoa

We would like to add forecasting of items for inventory purposes in an effort to make the app more full featured and use the machine learning capabilities to the fullest. Furthermore, we would like to improve the UX of the application by making it more performant in terms of the model and user engagement.