🎼 AI Symphony

Orchestrate AI instruction pipelines with elegance and power

AI Symphony is a powerful orchestration framework for chaining AI-powered instructions with seamless variable substitution and multi-model support. Create sophisticated AI workflows that combine multiple processing steps into cohesive, intelligent pipelines.

✨ Key Features

• 🤖 Multi-Model Support - Works with Gemini, Claude, OpenAI, Ollama, and Groq
• 🔗 Seamless Handshake - Automatic variable passing between pipeline steps
• 📝 Uniform Syntax - Consistent {{variable}} template format across all instructions
• 🎯 Flexible Mapping - Custom variable mapping between steps without template modification
• 🛡️ Error Recovery - Robust error handling and pipeline continuation options
• 📋 YAML Configuration - Declarative pipeline definitions for complex workflows
• 🔍 Template Validation - Built-in validation for template variables and pipeline configuration
• 📊 Progress Tracking - Real-time pipeline execution monitoring and metrics

🚀 Quick Start

Installation

pip install ai-symphony

Basic Usage

from ai_symphony import AISymphony

# Initialize with your instruction templates directory
symphony = AISymphony(instruction_dir="ai-instruction")

# Run a single instruction
result = symphony.run_instruction(
    instruction_file="analyze_text.j2",
    variables={"input": "Your text here", "focus": "main_points"}
)

print(result)

Pipeline Example

# Define a multi-step pipeline
steps = [
    {
        "name": "extract",
        "instruction_file": "extract_facts.j2",
        "temperature": 0.1
    },
    {
        "name": "analyze", 
        "instruction_file": "analyze_facts.j2",
        "variables": {"analysis_type": "comprehensive"},
        "temperature": 0.3
    },
    {
        "name": "summarize",
        "instruction_file": "create_summary.j2",
        "variables": {"max_sentences": "3"}
    }
]

# Execute the pipeline
result = symphony.execute_pipeline(
    steps=steps,
    initial_input="Your document content...",
    pipeline_name="document_analysis"
)

if result.success:
    print("Final output:", result.final_output)
    print(f"Completed in {result.total_duration_seconds:.1f} seconds")

YAML Configuration

# pipeline.yaml
name: "content_processor"
description: "Process and enhance content through AI pipeline"
continue_on_error: false
enable_handshake: true

steps:
  - name: "extract_key_points"
    instruction_file: "extract_points.j2"
    temperature: 0.2

  - name: "enhance_content"
    instruction_file: "enhance_writing.j2"
    variables:
      style: "professional"
      tone: "engaging"
    temperature: 0.4

# Execute YAML pipeline
result = symphony.execute_pipeline_from_yaml("pipeline.yaml")

🎯 Use Cases

📝 Content Processing

• Document summarization and analysis
• Language translation and localization
• Writing enhancement and style transformation
• Content generation and editing workflows

🔍 Data Analysis

• Sentiment analysis and classification
• Information extraction from unstructured data
• Pattern recognition and insight generation
• Multi-step analytical workflows

🤖 AI Workflow Automation

• Multi-model AI chains for complex tasks
• Fallback strategies and error recovery
• Dynamic pipeline generation
• Batch processing with consistent quality

🏢 Enterprise Integration

• Document processing pipelines
• Customer support automation
• Content moderation workflows
• Research and analysis automation

🏗️ Architecture

AI Symphony is built around several core components:

• 🎼 AISymphony Core - Main orchestrator coordinating all components
• 📝 Composer - Template engine with uniform variable substitution
• 🔄 Pipeline Executor - Manages multi-step pipeline execution with handshake mechanisms
• 🤝 Handshake Manager - Enables flexible variable mapping between steps
• 🔌 Client System - Unified interface for multiple AI providers
• ✅ Validation System - Template and pipeline configuration validation

📚 Documentation

• 📖 Quick Start Guide - Get up and running in minutes
• 🏗️ Architecture Overview - Understand how AI Symphony works
• 🎯 Examples Collection - Comprehensive examples and case studies
• 📋 API Reference - Detailed API documentation

🌟 Examples

Explore our comprehensive example collection:

• 🚀 Basic Examples - Learn the fundamentals
• 📝 Text Processing - Document manipulation workflows
• 🔍 Analysis & Insights - Extract insights from data
• 🔄 Multi-Model Workflows - Leverage multiple AI providers
• ⚡ Advanced Techniques - Push the boundaries

🛠️ Supported AI Providers

Provider	Status	Models
Google Gemini	✅	Gemini Pro, Gemini Flash
Anthropic Claude	✅	Claude 3 Sonnet, Claude 3 Haiku
OpenAI	✅	GPT-4, GPT-3.5 Turbo
Ollama	✅	Llama 2, Mistral, CodeLlama
Groq	✅	Llama 2 70B, Mixtral 8x7B

⚙️ Configuration

Set up your AI provider API keys:

# Choose one or more providers
export GEMINI_API_KEY="your-gemini-key"
export ANTHROPIC_API_KEY="your-claude-key"  
export OPENAI_API_KEY="your-openai-key"
export GROQ_API_KEY="your-groq-key"
# Ollama runs locally, no API key needed

🤝 Contributing

We welcome contributions! Please see our Contributing Guide for details.

Development Setup

# the repository is not published; the code lives in ~/ai-mac-studio/agents/infrastructure/ai-symphony
cd ai-symphony
pip install -e ".[dev]"

Running Tests

python test_ai_symphony.py

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• Built with ❤️ by Gerd Graßhoff
• Inspired by the need for elegant AI workflow orchestration
• Thanks to all contributors and the AI community

🔗 Links

• 📦 PyPI Package

---

Ready to orchestrate your AI workflows? Start with our Quick Start Guide and join the AI Symphony! 🎼