Skip to content

GETTING_STARTED.md

Latest commit

 

History

History
303 lines (226 loc) · 8.09 KB

GETTING_STARTED.md

File metadata and controls

303 lines (226 loc) · 8.09 KB

🚀 Getting Started with AI Council

Welcome to AI Council! This guide will help you get up and running quickly with the most powerful multi-agent AI orchestration system.

📋 Prerequisites

  • Python 3.8+ (Python 3.10+ recommended)
  • Git for cloning the repository
  • Basic Python knowledge for integration

🔧 Installation

1. Clone the Repository

git clone https://github.com/yourusername/ai-council.git
cd ai-council

2. Install Dependencies

# Install in development mode
pip install -e .

# Or install with all development dependencies
pip install -e ".[dev]"

3. Verify Installation

# Run the validation script
python scripts/validate_infrastructure.py

# Run basic tests
python -m pytest tests/test_core_models.py -v

You should see:

✅ ALL INFRASTRUCTURE COMPONENTS VALIDATED SUCCESSFULLY!
✅ Ready for orchestration component implementation

🎯 Your First AI Council Request

Let's start with a simple example:

1. Create Your First Script

Create a file called my_first_ai_council.py:

#!/usr/bin/env python3
"""
My First AI Council Example
"""

from ai_council.factory import AICouncilFactory
from ai_council.core.models import ExecutionMode

def main():
    print("🚀 Initializing AI Council...")
    
    # Create AI Council instance
    factory = AICouncilFactory()
    ai_council = factory.create_ai_council_sync()
    
    # Your first request
    question = "What are the main benefits of renewable energy?"
    
    print(f"❓ Question: {question}")
    print("🔄 Processing...")
    
    # Process the request
    response = ai_council.process_request_sync(
        question,
        ExecutionMode.BALANCED
    )
    
    # Display results
    print("\n" + "="*50)
    print("📝 AI Council Response:")
    print("="*50)
    print(response.content)
    print(f"\n📊 Confidence: {response.overall_confidence:.2f}")
    print(f"🤖 Models Used: {', '.join(response.models_used)}")
    print(f"💰 Cost: ${response.cost_breakdown.total_cost:.4f}")
    print("="*50)

if __name__ == "__main__":
    main()

2. Run Your First Script

# Set Python path (Windows)
$env:PYTHONPATH = "."

# Run your script
python my_first_ai_council.py

You should see output like:

🚀 Initializing AI Council...
❓ Question: What are the main benefits of renewable energy?
🔄 Processing...

==================================================
📝 AI Council Response:
==================================================
Renewable energy adoption offers significant environmental benefits and long-term economic advantages, though initial costs and infrastructure challenges must be considered.

📊 Confidence: 0.88
🤖 Models Used: model-1, model-2, model-3
💰 Cost: $0.0210
==================================================

🎮 Try Different Execution Modes

Now let's explore the different execution modes:

#!/usr/bin/env python3
"""
Exploring AI Council Execution Modes
"""

from ai_council.factory import AICouncilFactory
from ai_council.core.models import ExecutionMode

def compare_execution_modes():
    factory = AICouncilFactory()
    ai_council = factory.create_ai_council_sync()
    
    question = "Explain machine learning in simple terms"
    
    modes = [
        (ExecutionMode.FAST, "🚀 FAST"),
        (ExecutionMode.BALANCED, "⚖️ BALANCED"), 
        (ExecutionMode.BEST_QUALITY, "💎 BEST_QUALITY")
    ]
    
    print(f"Question: {question}\n")
    
    for mode, mode_name in modes:
        print(f"--- {mode_name} MODE ---")
        
        # Get cost estimate first
        estimate = ai_council.estimate_cost_and_time(question, mode)
        print(f"💰 Estimated Cost: ${estimate.total_cost:.4f}")
        print(f"⏱️  Estimated Time: {estimate.total_time:.1f}s")
        
        # Process request
        response = ai_council.process_request_sync(question, mode)
        
        print(f"📊 Confidence: {response.overall_confidence:.2f}")
        print(f"💵 Actual Cost: ${response.cost_breakdown.total_cost:.4f}")
        print(f"📝 Response: {response.content[:100]}...")
        print()

if __name__ == "__main__":
    compare_execution_modes()

🔍 Understanding the Results

When AI Council processes your request, you get:

Response Object

  • content: The final synthesized answer
  • overall_confidence: How confident the system is (0.0 to 1.0)
  • models_used: Which AI models were involved
  • cost_breakdown: Detailed cost information
  • execution_metadata: Processing details

Confidence Scores

  • 0.9-1.0: Very high confidence, reliable answer
  • 0.7-0.9: Good confidence, generally trustworthy
  • 0.5-0.7: Moderate confidence, consider verification
  • 0.0-0.5: Low confidence, requires careful review

🛠️ Next Steps

1. Explore Examples

# Run the comprehensive examples
python examples/basic_usage.py
python examples/complete_integration.py
python examples/orchestration_example.py

2. Try Different Task Types

AI Council handles various task types:

# Code generation
response = ai_council.process_request_sync(
    "Write a Python function to calculate fibonacci numbers",
    ExecutionMode.BALANCED
)

# Research and analysis  
response = ai_council.process_request_sync(
    "Research the latest developments in quantum computing",
    ExecutionMode.BEST_QUALITY
)

# Creative writing
response = ai_council.process_request_sync(
    "Write a short story about AI and humans working together",
    ExecutionMode.BALANCED
)

3. Learn About Configuration

from ai_council.utils.config import load_config

# Load and examine the configuration
config = load_config()
print(f"Available models: {list(config.models.keys())}")
print(f"Execution modes: {list(config.execution_modes.keys())}")

4. Explore Advanced Features

  • Custom Configuration: Create your own execution modes and routing rules
  • Cost Optimization: Fine-tune cost vs. quality trade-offs
  • System Monitoring: Monitor system health and performance
  • Batch Processing: Process multiple requests efficiently

📚 Learning Resources

Documentation

Examples

🆘 Troubleshooting

Common Issues

"Module not found" errors

# Make sure Python path is set
$env:PYTHONPATH = "."  # Windows
export PYTHONPATH="."  # Linux/Mac

Tests failing

# Run the validation script
python scripts/validate_infrastructure.py

# Check specific test
python -m pytest tests/test_core_models.py -v

Configuration issues

# Check configuration loading
python -c "from ai_council.utils.config import load_config; print(load_config())"

Getting Help

  1. Check the documentation in the docs/ directory
  2. Run the examples to see working code
  3. Check the test files for usage patterns
  4. Review the system validation report for system status

🎉 Congratulations!

You've successfully set up AI Council and made your first request! You now have access to a production-grade multi-agent AI orchestration system.

What You've Learned

  • ✅ How to install and set up AI Council
  • ✅ How to make your first AI request
  • ✅ Understanding execution modes and their trade-offs
  • ✅ How to interpret AI Council responses
  • ✅ Where to find more advanced features

Next Steps

  • Explore the comprehensive examples
  • Read the architecture documentation
  • Try custom configurations
  • Integrate AI Council into your projects

Welcome to the future of AI orchestration! 🚀