NCERTMate
An AI-assisted learning companion for explaining NCERT concepts in clearer, more student-friendly language.
What it is
NCERTMate is a Vite, React, and TypeScript learning experiment for making NCERT explanations clearer and easier to follow. It uses the Gemini API to re-explain textbook concepts in plainer, more student-friendly language.
Why I built it
I wanted textbook learning to feel less like memorisation and more like having someone patiently explain the idea until it clicks.
What problem it solves
Students hit a concept, search for an easier explanation, and often lose the thread. NCERTMate keeps the explanation close to the source material, so the learning path stays continuous and book-grounded.
For learning tools, the hard part is not just answering. It is staying close to the source material, explaining at the right level, and being honest about uncertainty.
Stack
- Vite + React for a fast, content-first SPA.
- TypeScript throughout.
- Gemini API for explanation and concept scaffolding.
Current status
Lab. An experiment, not a finished product. The live site is a working surface for the idea, not a polished release.
What I learned
- “Explain” is a harder prompt target than “summarise”, and far more useful.
- For learning tools, staying close to the source text is what makes the output trustworthy.
- The interface between curriculum structure and AI output is where the real design work lives.
What I would improve next
- More granular concept-graph navigation (chapter, concept, prerequisite).
- A self-test mode that generates questions from the explanation.
- Tighter citation back to the source textbook page.
Links
- Live: ncertmate.vercel.app