[FEATURE] Refine Quiz system component

[NehaP1706]

Feature Description

Currently, the QUIZ component involves 10 questions per course in all semesters. I propose using a database to store questions instead of hard-coding them onto the website. This allows community contribution, a larger question-set and true learning nature as the answers won't be publicly visible.

Problem It Solves

It improves on the current status of the feature and brings it closer to a production-level build. A student could previously navigate to the repository and view the answers, defeating the purpose of the testing setup.

Proposed Solution

A (Prisma) database connection shall be set up in lib\ instead of the current question list. Current consideration of the file structure is:

lib/
  quizData.ts        (Keep types only (Quiz, Question interfaces))
  db.ts              (Database connection (Prisma))
app/
  quiz/
    page.tsx         (Fetch subjects from DB (server component))
    [slug]/
      page.tsx      (Fetch questions from DB (server component))
      QuizClient.tsx (No changes needed (receives Quiz prop as before))
prisma/
  schema.prisma      (DB schema)

A swap-in ready AI version shall be kept in comments for further improvement in another issue. Discussion on this aspect is necessary - design and architecture choices for latency and accuracy.

Additional Context

• Limited changes are expected in about 3 files on the system.
• Due to an extended dependency, package-lock.json etc shall change (prisma).
• A seeding script will be made available for initial startups.

Kindly assign this issue to me under GSSOC'26. Further discussion is encouraged in this thread.

[NehaP1706]

I'm ready with all the required adjustments. The PR will include a detailed guide on how to set up and seed the database. Please let me know as soon as possible so we can accelerate this stage!

[pushkarscripts]

Thanks for the detailed proposal. The overall direction and architecture considerations make sense, and I do agree that scalability and maintainability of the quiz system will become important as the platform grows.

However, introducing Prisma/database infrastructure right now would also increase the overall project complexity, deployment requirements, contributor setup complexity, and maintenance overhead, especially while multiple other features and structural changes are actively being worked on during GSSoC.

For now, I would prefer keeping the quiz system lightweight and filesystem/static-data driven until the current structure stabilizes a bit more.

This is definitely something that can be revisited during the later stages of GSSoC once the current quiz system and other ongoing changes settle properly.

That said, the planning and thought process here is genuinely good.

[NehaP1706]

Alright, I assume introducing AI (mostly to reduce the repetitiveness of the quiz) to the generation component shall also be on hold?

[pushkarscripts]

Alright, I assume introducing AI (mostly to reduce the repetitiveness of the quiz) to the generation component shall also be on hold?

One thing that could still work well for now is introducing a lightweight randomization system where the quiz dynamically selects 10 questions out of a larger pool of, say, 25-50 questions for a subject/topic.

Topic-wise or module-wise quizzes similar to W3Schools would also make sense at the current stage and would already help reduce repetition significantly without needing AI/API integrations yet.

Anything involving API credits/costs is probably not very sustainable right now since the platform traffic is still relatively low and I currently do not have enough financial routes to support those costs properly in the long run.

So for now, I’d prefer keeping the system lightweight, static-data driven, and easy to maintain while still improving the quiz experience incrementally.