ain shams

A full-stack AI research tool that turns any Wikipedia topic into study material.

The user enters a topic and selects an audience. The app fetches the live Wikipedia article, then generates bullet-point summaries, key vocabulary, and a multiple-choice quiz using a large language model — all grounded strictly in the retrieved source text.

The standout feature is an audience persona router. The same article produces two very different outputs depending on whether "Beginner" or "Academic" mode is selected, by dynamically adjusting both the model's temperature and the style instructions injected into each prompt.

It also calculates reading metrics (word count, reading time, and a Flesch readability score) in pure Python without any AI calls, suggests related articles by filtering the page's outgoing links, and exports everything as a downloadable Markdown study deck.

Built with a FastAPI backend using Pydantic validation, LangChain for tooling and prompt chaining, Groq for fast model inference, and a vanilla JavaScript frontend with no framework dependencies.