Product Requirements Document (PRD) for "India's OS for Learning" MVP

1. Project Overview This PRD outlines the requirements for a student-centric MVP of "India's OS for Learning." The platform is designed to be a personal AI-powered study companion that helps students efficiently manage and master their academic material. The focus is on providing tools that directly address common student pain points: disorganized notes, tedious rote memorization, and the need for self-assessment and personalized feedback. 2. Goals of the MVP Provide a Seamless Study Workflow: Create a continuous digital workflow from unstructured notes to actionable study tools. Demonstrate AI as a Value-Add: Showcase how AI can automate time-consuming tasks like note-taking and content creation. Establish a Conversion Funnel: Use core free features (Notes, Flashcards, Quizzes) to demonstrate value and encourage conversion to the premium Subjective Copy Check feature. Build Foundational Technology: Develop a scalable and robust backend on GCP that can handle future growth and new features. 3. User Personas & Stories Dedicated Student (The Primary User): As a student, I want to upload my lecture notes, textbook PDFs, or handwritten notes so I can have a single, organized digital library. As a student, I want to automatically generate flashcards and quizzes from my digital notes to quickly test my recall and understanding. As a student, I want to practice writing subjective answers for my exams and get an immediate, detailed evaluation on my strengths and weaknesses. As a student, I want to see a history of my test scores and performance analytics to track my progress and identify areas for improvement. 4. Core Features & Functionality (Developer Scope) The developer's role is to build the backend and APIs that power the following student-facing features. The UI/UX will be handled by a separate design team, but the backend must support the functionality. a) Document & Notes Generation: Purpose: To convert unstructured study material into organized, digital notes. Backend Requirements: An API endpoint to receive and securely store uploaded files (PDFs, images) in Cloud Storage. A background process (e.g., Cloud Functions) to trigger an OCR and text extraction service (e.g., Document AI) on new uploads. Use a language model (on Vertex AI) to structure the raw text, identify key headings, and create a clean, formatted digital note. Store the processed notes and their metadata in a database (Firestore/Cloud SQL). API Endpoint: POST /api/documents/upload b) Flashcard Generation: Purpose: To automate the creation of study aids for memorization. Backend Requirements: An API endpoint that takes a document ID as input. The backend retrieves the notes and sends them to a Gen AI model (on Vertex AI) with a prompt to generate flashcards (front: term, back: definition) in a structured format (e.g., JSON). Save the generated flashcard data to the database, linked to the user's account and the source notes. API Endpoint: POST /api/documents/{documentId}/generate-flashcards c) Quiz Generation & Test Engine: Purpose: To provide a tool for self-assessment. Backend Requirements: An API endpoint to generate objective quizzes (multiple choice, true/false) from a document's content using a Gen AI model. An API to serve questions from the quiz/test bank. An API endpoint to receive user answers and provide instant, accurate feedback for objective questions. API Endpoints: POST /api/documents/{documentId}/generate-quiz, GET /api/quizzes/{quizId}/questions, POST /api/quizzes/{quizId}/submit d) Subjective Copy Check: Purpose: To provide personalized, detailed feedback on subjective answers, simulating a human evaluator. Backend Requirements: An API to accept a user's uploaded answer (text or image) and the original question. The backend sends the user's answer and the question to a powerful Gen AI model (e.g., a fine-tuned model on Vertex AI) with a detailed prompt. The prompt should instruct the model to: Evaluate the answer against a pre-defined set of criteria (e.g., clarity, structure, argument strength, grammar, relevance). Provide a numeric score out of 10. Give a written critique with specific points of improvement and suggestions. The feedback and score are returned to the user and saved in their profile history. API Endpoint: POST /api/tests/subjective-answer/evaluate 5. User Flow User lands on the platform and is immediately prompted to upload their first file to "See the magic." The file is uploaded to the backend and processed asynchronously. The user is redirected to a simple sign-up page (using Firebase Authentication or Cloud Identity). After signing up, the user is taken to their dashboard, where the newly processed notes are ready. From the dashboard, the user can click buttons to generate flashcards or quizzes from the notes. The user can navigate to the "Tests" section to upload their subjective answers for the copy-checking feature. Throughout their usage, the user's progress and activity are tracked to provide a personalized experience. 6. Technical Stack Cloud Platform: Google Cloud Platform (GCP) Core Services: Cloud Run (for stateless API services) and Cloud Functions (for event-driven triggers). AI/ML: Vertex AI (for model deployment), Document AI (for OCR and text extraction). Database: Cloud Firestore or Cloud SQL for structured data storage. File Storage: Cloud Storage. Authentication: Firebase Authentication. 7. Success Metrics Number of Documents Uploaded: A key indicator of user adoption. Feature Conversion Rate: Percentage of users who generate at least one flashcard deck or quiz. Subjective Copy Check Usage: Number of subjective answers submitted for evaluation, as this is a premium feature. User Retention: Percentage of users who return after one week.

Project ID: 39737703