Full Stack AI Developer needed for AI Powered Personalized Education Platform

We need personalised AI tutoring platform that provides students with a fully personalized learning experience. The platform must adapt to each student's current knowledge level, generate structured study plans, test their understanding, and provide on-demand academic help. The system must require no account creation — students should be able to open the app and start learning immediately. --- 2. Functional Requirements 2.1 Lesson Planner - The student must be able to enter a subject, list of topics, learning goal, available study hours per day, and number of days - Before generating a plan, the system must run a diagnostic quiz (4–6 questions) to assess the student's current knowledge level - Based on quiz results, the system must generate a personalized day-by-day study plan tailored to the student's skill gaps - Each day in the plan must include: activities, resources, learning objectives, and estimated duration - The plan must include weekly goals, study tips, and a skill summary - The student must be able to view all previously generated plans and switch between them - The system must allow the student to regenerate and improve an existing plan based on their assignment performance data 2.2 AI Assignments - The student must be able to generate an assignment for any subject and topic - The student must be able to select difficulty level (Easy / Medium / Hard) and number of questions (3–15) - If the student has a lesson plan, the system must suggest their plan subjects and auto-set difficulty based on their assessed skill level - The assignment must contain a mix of question types: Multiple Choice (MCQ), Short Answer, and Long Answer - On submission, the system must grade each answer and provide: - Score per question - Correct answers for wrong responses - Detailed feedback - An overall grade and summary - For every answer the student got correct, the system must generate a counter-question to verify genuine understanding (not guessing) - The student must be able to submit counter-question answers, which the system evaluates separately - All past assignments must be accessible from a history panel 2.3 Doubt Solver (AI Tutor Chat) - The student must be able to start an AI tutoring chat session - Before chatting, the student must be able to link the session to one of: - A specific assignment (the AI will have access to the questions, student answers, and what they got wrong) - A specific lesson plan (the AI will have access to the topics and plan content) - Neither (open free chat) - The AI must provide context-aware responses — explanations must be tailored to the student's assignment results or plan topics, not generic - The system must suggest relevant questions based on the student's linked context to help them get started - The student must be able to have a full back-and-forth conversation with follow-up questions 2.4 Study Resources - When a lesson plan is created, the system must automatically generate study material for each topic in the plan - Each topic's resource must include: explanation, key concepts, worked examples, and a quick-check question - Resources must be organized by topic and accessible from a dedicated Resources section - The student must be able to browse all generated resources and switch between them --- 3. Non-Functional Requirements - The platform must be responsive and usable on desktop, tablet, and mobile - The platform must support dark and light mode with user preference saved locally - All AI responses must be generated in real-time; the UI must show loading states during generation - The system must handle API errors gracefully and display user-friendly error messages --- 4. Technical Requirements - Frontend: Modern web framework (Next.js), deployed as a web app accessible via browser - Backend: REST API server handling all business logic and AI orchestration - Database: Relational database to persist lesson plans, assignments, submissions, and conversations - AI Model: OpenAI model for all content generation (quizzes, plans, assignments, grading, tutoring) - RAG Pipeline: The doubt solver must use Retrieval-Augmented Generation — the AI must retrieve relevant context from the student's linked assignment or lesson plan before responding