Skip to content

README.md

Latest commit

 

History

History
300 lines (235 loc) · 9.68 KB

README.md

File metadata and controls

300 lines (235 loc) · 9.68 KB

📚 Multi-Agent AI Study Assistant

A powerful, personalized learning platform powered by multiple specialized AI agents. Get custom learning roadmaps, practice quizzes, AI tutoring, and RAG-powered document Q&A to supercharge your studies!

🌟 Features

🎯 Personalized Learning Analysis

  • Student Profiling: AI analyzes your knowledge level, goals, and learning style
  • Gap Analysis: Identifies what you need to learn and prerequisite knowledge
  • Custom Recommendations: Tailored approach based on your unique needs

🗺️ Smart Learning Roadmaps

  • Structured Learning Paths: Break down complex topics into manageable phases
  • Time-Optimized: Plans adapted to your available study time
  • Milestone Tracking: Clear checkpoints to measure progress
  • Flexible Scheduling: Adjust pace based on your schedule

📝 Dynamic Quiz Generation

  • Adaptive Difficulty: Questions matched to your knowledge level
  • Multiple Question Types: MCQ, True/False, Short Answer, Problem-Solving
  • Detailed Explanations: Learn from both correct and incorrect answers
  • Custom Focus Areas: Target specific topics you want to practice

🤖 AI Tutor

  • 24/7 Availability: Get help whenever you need it
  • Personalized Explanations: Adapted to your learning style
  • Step-by-Step Guidance: Break down complex concepts
  • Real-World Examples: Understand through practical applications

📄 RAG-Powered Document Q&A

  • Upload Study Materials: PDFs, textbooks, lecture notes
  • Intelligent Search: Find relevant information instantly
  • Context-Aware Answers: Get answers grounded in your materials
  • Source Citations: Know where information comes from

🔍 Resource Finder

  • Curated Resources: AI finds the best learning materials
  • Quality Filtered: Only high-quality, relevant resources
  • Learning Style Matched: Resources suited to how you learn best

🏗️ Architecture

This project uses a multi-agent architecture where specialized AI agents collaborate:

study_agents.py          # Specialized AI agents (Analyzer, roadmap Creator, Quiz Generator, Tutor, Resource Finder, RAG Tutor)
agent_handler.py         # Orchestrates agent workflows and manages state
rag_helper.py            # RAG functionality for document-based learning
config.py                # Configuration management
prompts.yaml             # Agent personas and prompt templates
app.py                   # Streamlit web interface

Agent Roles

  1. Student Analyzer Agent: Assesses learning needs, identifies gaps, recommends approaches
  2. Roadmap Creator Agent: Designs personalized learning paths with phases and milestones
  3. Quiz Generator Agent: Creates adaptive assessments with explanations
  4. Tutor Agent: Provides explanations and answers questions
  5. Resource Finder Agent: Searches and recommends learning materials
  6. RAG Tutor Agent: Answers questions using uploaded study documents

🛠️ Technologies Used

  • Phidata: Multi-agent orchestration framework
  • Streamlit: Web interface
  • LangChain: Document processing and RAG
  • ChromaDB: Vector database for RAG
  • OpenAI/Groq: LLM providers
  • DuckDuckGo: Web search for resource finding

📝 Tips for Best Results

  1. Be Specific: The more detailed your learning goals, the better the roadmap
  2. Honest Assessment: Accurately rate your knowledge level for best recommendations
  3. Upload Materials: Use the RAG feature with your textbooks and notes for personalized help
  4. Regular Quizzes: Test yourself frequently to reinforce learning
  5. Follow the Roadmap: Trust the AI's structured approach
  6. Ask Questions: Use the tutor liberally - there are no dumb questions!

🎨 Learning Styles Explained

The AI adapts to your preferred learning style to maximize effectiveness. Choose the one that best describes how you learn:

👁️ Visual Learners

How you learn best: Through seeing and observing

Characteristics:

  • Prefer diagrams, charts, and images
  • Remember faces better than names
  • Like color-coded notes and mind maps
  • Benefit from watching demonstrations

Recommended resources:

  • 📊 Infographics and flowcharts
  • 🎥 Video tutorials and demonstrations
  • 🗺️ Mind maps and concept diagrams
  • 📈 Visual data representations
  • 🎨 Color-coded study materials

Study tips:

  • Use highlighters and color coding
  • Draw diagrams to understand concepts
  • Watch video explanations
  • Create visual summaries

🎧 Auditory Learners

How you learn best: Through listening and speaking

Characteristics:

  • Prefer verbal instructions
  • Remember what you hear
  • Enjoy discussions and lectures
  • Often talk through problems

Recommended resources:

  • 🎙️ Podcasts and audio lectures
  • 💬 Discussion forums and study groups
  • 🗣️ Verbal explanations and debates
  • 🎵 Educational songs or mnemonics
  • 📻 Audio books and recordings

Study tips:

  • Read notes aloud
  • Record and listen to lectures
  • Discuss topics with others
  • Use verbal repetition
  • Explain concepts out loud

🤸 Kinesthetic Learners

How you learn best: Through doing and hands-on experience

Characteristics:

  • Learn by doing and practicing
  • Prefer physical activity and movement
  • Like real-world applications
  • Need to try things yourself

Recommended resources:

  • 💻 Interactive coding platforms (for programming)
  • 🔬 Lab experiments and simulations
  • 🎮 Educational games and interactive tools
  • 🛠️ Hands-on projects and exercises
  • 🏃 Physical models and manipulatives

Study tips:

  • Build projects while learning
  • Use interactive simulations
  • Take frequent breaks to move
  • Practice with real examples
  • Create physical models or demonstrations

📖 Reading/Writing Learners

How you learn best: Through reading and writing

Characteristics:

  • Prefer written information
  • Love taking detailed notes
  • Enjoy reading textbooks and articles
  • Learn by writing summaries

Recommended resources:

  • 📚 Textbooks and comprehensive guides
  • 📝 Written tutorials and documentation
  • ✍️ Note-taking and summary exercises
  • 📄 Articles and research papers
  • 📋 Lists, definitions, and glossaries

Study tips:

  • Take extensive notes
  • Rewrite information in your own words
  • Create written summaries
  • Use flashcards with written content
  • Read and re-read materials

💡 How the AI Uses Your Learning Style

When you select your learning style, the AI agents will:

  1. Roadmap Creator: Structures your learning path with style-appropriate milestones
  2. Resource Finder: Prioritizes resources matching your preferred format
  3. Tutor Agent: Adapts explanations using style-specific techniques
  4. Quiz Generator: Includes questions that align with your learning preferences

Not sure which style fits you? Try the Mixed/Multimodal approach, which combines elements from all styles!

🐛 Troubleshooting

ModuleNotFoundError: No module named 'X'

Solution: Install missing dependencies

pip install -r requirements.txt

Common missing modules and their fixes:

  • phidata → pip install phidata
  • openai → pip install openai
  • chromadb → pip install chromadb
  • langchain_chroma → pip install langchain-chroma
  • langchain_openai → pip install langchain-openai
  • duckduckgo_search → pip install duckduckgo-search

"streamlit: command not found"

Solution: Use Python module syntax

python3 -m streamlit run app.py

Or add ~/.local/bin to your PATH:

export PATH="$HOME/.local/bin:$PATH"

"No API key found" or API errors

Solution:

  1. Ensure .env file exists in the project root
  2. Verify your API key is correctly formatted:
    • Groq keys start with gsk_
    • OpenAI keys start with sk-
  3. Check that the key is valid and active
  4. For OpenAI: Ensure you have credits in your account

"Failed to load document" (RAG feature)

Solution:

  • Ensure PDF is not password-protected
  • Check file is not corrupted
  • Try converting to text format first
  • Verify file size is reasonable (< 50MB recommended)

"Agent not responding" or timeout errors

Solution:

  • Check your internet connection
  • Verify API key is valid and active
  • Try switching to a different model in the app settings
  • For Groq: Check status.groq.com
  • For OpenAI: Check status.openai.com

ChromaDB errors

Solution:

# Delete the vector database and restart
rm -rf chroma_db/
python3 -m streamlit run app.py

Ensure you have write permissions:

chmod -R 755 .

Port already in use (Address already in use)

Solution:

# Kill existing Streamlit processes
pkill -f streamlit

# Or use a different port
streamlit run app.py --server.port 8502

🚀 Future Enhancements

  • Progress tracking and analytics
  • Spaced repetition scheduling
  • Collaborative study groups
  • Integration with learning platforms (Coursera, Udemy, etc.)
  • Mobile app version
  • Voice interaction for auditory learners
  • Gamification and achievements
  • Export to Anki flashcards

📄 License

This project is open source and available for educational purposes.

🙏 Acknowledgments

  • Built with Phidata multi-agent framework
  • Powered by Groq and OpenAI LLMs
  • Uses LangChain for document processing and RAG functionality

💬 Support

For issues, questions, or suggestions:

  1. Check the troubleshooting section above
  2. Review the configuration in prompts.yaml
  3. Ensure all dependencies are correctly installed

Happy Learning! 📚✨

Remember: The best way to learn is to start. Let AI agents guide your journey!