MemoryVerse AI

Design Decisions & Architecture MemoryVerse AI uses a decoupled architecture to ensure the UI remains fully responsive during heavy document processing. • Frontend: Built with React, TailwindCSS, and Framer Motion to handle the drag-and-drop ingestion and real-time dashboard updates. • Backend: A FastAPI (Python) server handles asynchronous API calls, preventing browser freezing. • AI Engine: We utilized Google Gemini 3.5 Flash via the google-genai SDK. It was chosen specifically for its high-speed inference and ability to return strictly constrained JSON, which is required to instantly map unstructured PDF data to the visual timeline. Setup & Testing To run this locally, clone the repository and follow the setup steps in the README.md. Note: Reviewers will need to supply their own Gemini API key by creating a `.env` file inside the `memoryverse-api` folder: GEMINI_API_KEY=your_actual_key_here Demo Please refer to the attached YouTube video link for a complete end-to-end demonstration of the automated skill extraction!