Amazon Purchases App

The app, designed to analyze Amazon purchases, handles a database exceeding 300 MB. To optimize performance, extensive data cleaning was performed, and the database was converted to the Parquet format, significantly improving loading times and efficiency. The app also integrates a language model (LLM) via Gemini, enabling personalized book recommendations based on users' profiles and purchase histories.

🥇 Winner of the Plotly App Challenge 2024 »

Available on Plotly Dash App Examples »

The app can be accessed here »

Award Description

Key Features

The app is divided into three main pages:

• Purchase overview: The application provides a detailed view of purchases from 2018 to 2022. Data can be analyzed year-over-year using an interactive filter, allowing users to explore and understand purchases trends.

• Customer demographics: In addition to purchases data, the application provides a demographic analysis of users, including gender, age, education level, household income, and geographic distribution by state.

• Book Recommendation Chatbot: Recognizing the importance of books as the primary product sold, a dedicated page for personalized book recommendations was created. These recommendations are generated by the Gemini language model, which analyzes each user's profile and book purchase history to offer highly relevant suggestions.

Overview

Built With

This project has been built using Python, Plotly Dash framework, CSS and Google Cloud Platform.

Contact

For feedback and notes, please reach out to Mayara Daher.