AI Tutoring Website Development

I’m launching an AI-driven tutoring site where visitors can pick absolutely any topic and study it their way—one-to-one, in a group room, or through self-paced lessons. Your job is to design and build the entire platform, weaving together a smooth user experience with the interactive tools that keep learners engaged.

Core experience
• Topic-agnostic structure: from the homepage a learner should be able to search or request any subject, then instantly see available tutors, group sessions, or self-paced tracks.
• Interactive learning: quizzes that auto-grade, a real-time live-chat window embedded in every lesson page, and discussion forums that sit under each course for peer support.
• Flexible course builder: I need an admin panel where new lessons, modules, and quizzes can be created without touching code, so the catalogue can grow to “any topic” on demand.
• AI layer: the site should surface recommended resources or next steps using an LLM (OpenAI API, LangChain or similar—open to your suggestion).

Key deliverables
1. Responsive front-end (React, Vue or similar) themed to a clean, modern academic look.
2. Scalable back-end (Node/Express, Django or comparable) with REST or GraphQL endpoints, user auth, and role-based dashboards for tutors, learners, and admins.
3. Quiz engine, WebSocket-powered live chat, and a forum module (integrated Discourse, Flarum, or a custom lightweight board).
4. Database schema, seed data, and migration scripts.
5. Deployment scripts (Docker or equivalent) plus step-by-step launch documentation.

Acceptance criteria
• A new user can register, enroll in a topic of their choosing, complete at least one quiz, post in a forum thread, and chat live—all without errors on mobile or desktop.
• Admin can create a fresh course, attach lessons and quizzes, and publish it in under five minutes.
• Page load times under three seconds for public pages and under five seconds for authenticated views.

If you’ve built LMS platforms, tutoring sites, or real-time apps before, I’d love to see your work. Let me know which stack you suggest and the rough timeline you’d need to deliver the MVP with the features above.