Turn PDFs and Videos into Smart Study Notes
Upload content and get structured notes, flashcards, quizzes, and mind maps. Transform your learning with AI-powered study tools.
Powerful Learning Features
Transform your study experience with our AI-powered learning tools
Smart Notes
AI-generated structured notes extracted from PDFs, videos, and other content formats.
Flashcards
Automatically generated flashcards to help you memorize key concepts effectively.
Quizzes
Test your knowledge with AI-generated quizzes based on your uploaded content.
Mind Maps
Visualize connections between concepts with interactive mind maps.
How Adhyan Works
A comprehensive guide to our AI-powered content transformation system
Discover how we process your learning materials, filter the content intelligently, and use Google's Gemini AI to create multiple tailored study formats
Content Upload & Type Detection
Upload your learning material in multiple formats.
- Supported formats: PDF documents, YouTube video URLs, and plain text.
- The content type is automatically detected through validation checks. YouTube links are verified using regular expressions, and all inputs undergo quality checks to ensure they meet minimum standards.
Source-Specific Extraction
Content is extracted using specialized methods for each format.
- PDF extraction uses unpdf to preserve structure; YouTube transcripts are retrieved using Tactiq API.
- Text normalization ensures sentence integrity, while text statistics (word/sentence count, token usage) guide further processing.
Content Filtering & Preparation
Content is filtered and organized for optimal learning.
- Content is chunked into 100–1000 characters and cleaned of redundancy or irrelevant info.
- Key concepts are identified, prioritized, and aligned to focus learning.
Gemini AI Structured Notes
Google's Gemini AI transforms filtered content into organized notes.
- Gemini 2.0 Flash processes content using specialized prompts and structures it into with sections, takeaways and summary.
- Definitions, examples, and links are categorized explicitly, guided by system prompts for optimal learning format.
Flashcard Generation
Three specialized flashcard types with targeted content extraction.
- Three distinct flashcard types are generated: Definition cards (term-definition pairs), Recall cards (key points and headings), and Application cards (examples and conceptual connections).
- Each type uses tailored data extraction, ensuring the content is filtered to best suit its learning purpose with custom Gemini prompts guiding how information is selected and structured.
Flashcard Types & Content Processing
Definition Cards
Directly parses term-definition pairs from 'definitions' arrays and formats them as 'What is X?' question cards.
Recall Cards
Uses LangChain to transform extracted points, headings, and key takeaways into question-answer pairs via Gemini AI.
Application Cards
Identifies examples and connections in content, extracts related concepts, and generates scenario-based cards with context.
Interactive Quiz System
AI-generated quizzes with adaptive difficulty and content prioritization.
- Two Quiz Types with Customization: Users can choose between MCQs and True/False quizzes, with options to set difficulty and number of questions for a tailored learning experience.
- Smart Content Extraction & Balancing: Quiz content is extracted and scored based on importance which ensures ensures high-quality, well-rounded questions.
Quiz Format & Content Extraction
MCQ Generation
Analyzes content importance based on type and scores them accordingly (definitions: 3x, examples: 2.5x, points: 2x) to generate 4-option questions.
True/False Creation
Generates balanced sets of true statements and carefully modified false statements with custom prompt instructions.
Content Distribution
Creates a balanced mix across content types: definitions (25%), concepts (50%), applications (25%).
Visual Mind Map Creation
Complex relationships are visualized through interactive mind maps.
- Gemini analyzes JSON notes to identify relationships and builds maps with a central concept and 3–6 main branches.
- Content is prioritized by importance, and cross-branch conceptual links are intelligently created.
Mind Map Node Types & Processing
Concept Nodes
Primary ideas marked with 'high' emphasis level, organized in radial branches from a central concept with specialized positioning algorithms.
Detail Nodes
Supporting facts and elaborations marked with 'medium' emphasis, arranged in optimal positions using mathematical angle calculations.
Example Nodes
Concrete illustrations with 'low' emphasis level, connected to parent concepts with color-coded edges based on node type.