Skip to main content

Feature Specification: AI Textbook Platform

Branch: 1-ai-textbook-platform | Date: 2025-12-10

Overview​

This feature entails the development of a comprehensive, interactive AI textbook platform aligned with the Panaversity outline. The platform will cover a 13-week curriculum focusing on advanced robotics and AI topics, including ROS 2, Gazebo, NVIDIA Isaac, VLA (Vision-Language-Action), and practical application with Jetson/Unitree hardware, suitable for both cloud and on-premise environments.

User Scenarios & Testing​

User Scenarios​

  1. Student Enrollment: A new student signs up for the platform, completes a personalization survey, and gains access to the course content.
  2. Content Access: A student navigates through the textbook chapters, including theory, diagrams, code examples, and interactive labs.
  3. Assessment Completion: A student takes a quiz (MCQs) or submits a capstone project for grading.
  4. RAG Chatbot Interaction: A student uses the RAG chatbot to ask questions about the textbook content and receives relevant, context-aware answers.
  5. Personalized Learning: The platform suggests relevant content or exercises based on the student's progress and performance.
  6. Urdu Content Access: An Urdu-speaking student switches the platform language to Roman Urdu to access translated content and chatbot responses.
  7. Instructor Content Management: An instructor (or content creator) adds, updates, or removes textbook chapters, assessments, and labs.
  8. Authentication Management: A user logs in, manages their profile, and updates personalization settings.

Acceptance Criteria​

  • Student Enrollment:
    • Given a new user wants to access the platform
    • When they complete the sign-up process via Better-Auth and fill out the personalization survey
    • Then they are granted access to the course materials and their personalization preferences are stored.
  • Content Access:
    • Given a student is logged in and navigates to a chapter
    • When they select a chapter
    • Then the chapter content (theory, diagrams, code, labs, MCQs) is displayed correctly, with interactive elements functioning as expected.
  • Assessment Completion:
    • Given a student is viewing an assessment
    • When they complete and submit an MCQ quiz or a capstone project
    • Then their submission is recorded, and feedback/results are available appropriately (e.g., immediate for MCQs, later for capstones).
  • RAG Chatbot Interaction:
    • Given a student is on any content page
    • When they activate the RAG chatbot and ask a question related to the textbook
    • Then the chatbot provides a concise, accurate, and contextually relevant answer sourced from the textbook content.
  • Personalized Learning:
    • Given a student has an established learning profile
    • When they interact with the platform over time
    • Then the platform identifies learning gaps or areas of interest and suggests content or tasks to address them.
  • Urdu Content Access:
    • Given an Urdu-speaking user is logged in
    • When they select "Roman Urdu" as their preferred language
    • Then all applicable textual content, including the RAG chatbot's responses, are rendered in Roman Urdu.
  • Instructor Content Management:
    • Given an authorized instructor is logged in
    • When they use the content management tools
    • Then they can effectively create, edit, publish, and unpublish textbook chapters, labs, and assessments.
  • Authentication Management:
    • Given a user is on the platform
    • When they attempt to log in, log out, or update their profile/personalization settings
    • Then these actions are securely processed by Better-Auth, and changes are reflected accurately.

Functional Requirements​

  1. Textbook Content Delivery: The platform shall display comprehensive textbook content including text, images, diagrams, code blocks, and embedded interactive elements (e.g., simulators, code editors).
  2. Assessment System: The platform shall provide multiple-choice questions (MCQs) and support submission/evaluation of capstone projects.
  3. User Authentication & Authorization: The platform shall implement a secure user authentication system using Better-Auth, supporting sign-up, login, and session management.
  4. User Profile & Personalization: The platform shall allow users to create and manage profiles, including personalization preferences based on an initial survey and ongoing learning activity.
  5. Multilingual Support (Urdu): The platform shall support content and UI translation to Roman Urdu.
  6. RAG Chatbot Integration: The platform shall integrate a RAG chatbot capable of answering user queries based on the textbook content, utilizing OpenAI SDK and Qdrant for vector search.
  7. Content Management System (CMS): The platform shall provide tools for instructors to create, edit, and publish textbook modules, chapters, and assessments.
  8. Deployment: The frontend shall be deployable on GitHub Pages, and the backend on Render, with a production-ready setup.
  9. Data Storage: The platform shall utilize Neon Postgres for relational data and Qdrant (free tier) for vector embeddings.

Success Criteria​

  1. Content Engagement: 80% of registered students complete at least 50% of the core curriculum within the 13-week period.
  2. Assessment Completion Rate: 90% of students attempt and submit at least 75% of the assigned assessments.
  3. RAG Chatbot Accuracy: The RAG chatbot provides relevant answers to content-related queries with 90% accuracy, as rated by users.
  4. Urdu Content Adoption: At least 20% of users in relevant regions utilize the Roman Urdu translation feature.
  5. Platform Availability: The platform (frontend and backend) maintains 99.9% uptime.
  6. Deployment Efficiency: New content updates can be deployed to production within 15 minutes.

Key Entities​

  • User: id, username, email, password_hash, profile_data (personalization preferences), language_preference.
  • Course: id, title, description, duration_weeks.
  • Chapter: id, course_id, title, content (MDX format), order.
  • Assessment: id, chapter_id (optional), type (MCQ, Capstone), title, data (questions/rubric).
  • Submission: id, assessment_id, user_id, submission_data, score, feedback.
  • VectorEmbedding: id, content_id, vector_data, source_text.
  • Translation: id, original_text_hash, language, translated_text.

Assumptions​

  1. The textbook content will be provided in a structured format suitable for MDX conversion.
  2. The RAG chatbot will primarily answer questions directly from the textbook content; complex conversational abilities are out of scope.
  3. "Production-ready" implies standard performance, security, and scalability for a learning platform, within the constraints of chosen free/low-cost tiers (Qdrant free, GitHub Pages, Render).
  4. Urdu translation will be for Roman Urdu and primarily for static content and RAG chatbot responses. Dynamic user-generated content might have limited translation.
  5. The capstone project evaluation might require manual instructor intervention, with the platform facilitating submission and grading records.

Open Questions / Clarifications​

[NEEDS CLARIFICATION: Specific metrics for "Production-ready" performance goals (e.g., latency, concurrent users)?] [NEEDS CLARIFICATION: Detailed scope of personalization features beyond initial survey (e.g., adaptive learning paths, spaced repetition)?] [NEEDS CLARIFICATION: Requirements for content versioning and collaboration for instructors?]