ai-flashcard-generator

AI Flashcard generator using Exabase

This is a simple example of an AI flashcard generator using Exabase, Next.js, AI SDK, shadcn/ui, and Tailwind CSS.

Users upload PDFs, the app chooses the most relevant documents for a prompt, and the AI generates flashcards that are saved back to Exabase memories in the base.

How Exabase is used in this example

• Store uploaded PDF files in the base.
• Extract and store a summary for each document so the LLM can choose which documents to use for flashcard generation.
• Save generated flashcards as memories for retrieval and reuse.

Demo

demo-x2.mp4

Deploy your own

Technologies

• Exabase: Headless cloud filesystem, automatic content extraction, memory generation and retrieval
• Next.js: Front-end and back-end
• AI SDK: AI capabilities (OpenAI model)
• shadcn/ui: Accessible UI components
• Tailwind CSS: Styling
• Biome: Lint and format

Run locally

bun install
bun run dev

Create a .env.local file with EXABASE_API_KEY and OPENAI_API_KEY. Optional variables are described in .env.example.

Open http://localhost:3000 and click New base to create an Exabase Base; you are redirected to /b/<baseId> where the app runs.