Advanced Research System (Based on Anthropic's Paper)

An enhanced implementation of the orchestrator-worker pattern from Anthropic's paper, "How we built our multi-agent research system", built on top of the bleeding-edge multi-agent framework swarms. Our implementation of this advanced research system leverages parallel execution, LLM-as-judge evaluation, and professional report generation with export capabilities.

📦 Installation

pip3 install -U advanced-research

# uv pip install -U advanced-research

Environment Setup

Create a .env file in your project root:

# Claude API Key (Primary LLM)
ANTHROPIC_API_KEY="your_anthropic_api_key_here"

# Exa Search API Key
EXA_API_KEY="your_exa_api_key_here"

# Optional: OpenAI API Key (alternative LLM)
OPENAI_API_KEY="your_openai_api_key_here"

🚀 Quick Start

Basic Usage

from advanced_research import AdvancedResearch

# Initialize the research system
research_system = AdvancedResearch(
    name="AI Research Team",
    description="Specialized AI research system",
    max_loops=1,
)

# Run research and get results
result = research_system.run(
    "What are the latest developments in quantum computing?"
)
print(result)

With Export Functionality

from advanced_research import AdvancedResearch

# Initialize with export enabled
research_system = AdvancedResearch(
    name="Quantum Computing Research",
    description="Research team focused on quantum computing advances",
    max_loops=1,
    export_on=True,  # Enable JSON export
)

# Run research - will automatically export to JSON file
research_system.run(
    "What are the latest developments in quantum computing?"
)
# Results will be saved to a timestamped JSON file

Advanced Configuration

from advanced_research import AdvancedResearch

# Initialize with custom settings
research_system = AdvancedResearch(
    name="Medical Research Team",
    description="Specialized medical research system",
    director_model_name="claude-3-5-sonnet-20250115",  # Use latest Claude model
    worker_model_name="claude-3-5-sonnet-20250115",
    director_max_tokens=10000,
    max_loops=2,  # Multiple research iterations
    output_type="all",  # Include full conversation history
    export_on=True,
)

# Run research with image input (if applicable)
result = research_system.run(
    "What are the most effective treatments for Type 2 diabetes?",
    img=None  # Optional image input
)

Batch Processing Multiple Queries

from advanced_research import AdvancedResearch

# Initialize the system
research_system = AdvancedResearch(
    name="Batch Research System",
    max_loops=1,
    export_on=True,
)

# Process multiple research tasks
tasks = [
    "Latest advances in renewable energy storage",
    "Current state of autonomous vehicle technology",
    "Recent breakthroughs in cancer immunotherapy"
]

# Run batch processing
research_system.batched_run(tasks)

Using Different Output Formats

from advanced_research import AdvancedResearch

# Initialize with specific output type
research_system = AdvancedResearch(
    name="Research System",
    output_type="json",  # Options: "all", "json", "markdown"
    export_on=False,  # Get results directly instead of exporting
)

# Run research and get formatted output
result = research_system.run(
    "What are the key challenges in AGI development?"
)

# Check available output methods
available_formats = research_system.get_output_methods()
print(f"Available output formats: {available_formats}")

📋 Quick Reference

Task	Code	Documentation
Basic Research	AdvancedResearch().run("query")	Basic Usage →
Export Results	AdvancedResearch(export_on=True)	Export Config →
Batch Processing	system.batched_run([queries])	Batch Processing →
Custom Models	AdvancedResearch(director_model_name="model")	Advanced Config →
Output Formats	AdvancedResearch(output_type="json")	Output Types →

✨ Key Features

Feature	Description
Orchestrator-Worker Architecture	A Director Agent coordinates research strategy while specialized worker agents execute focused search tasks with Exa API integration.
Advanced Web Search Integration	Utilizes exa_search with structured JSON responses, content summarization, and intelligent result extraction for comprehensive research.
High-Performance Parallel Execution	Leverages ThreadPoolExecutor to run multiple specialized agents concurrently, achieving significant time reduction for complex queries.
Flexible Configuration	Customizable model selection (Claude, GPT), token limits, loop counts, and output formatting options.
Conversation Management	Built-in conversation history tracking with the swarms framework's Conversation class for persistent dialogue management.
Export Functionality	JSON export with automatic timestamping, unique session IDs, and comprehensive conversation history.
Multiple Output Formats	Support for various output types including JSON, markdown, and full conversation history formatting.
Session Management	Unique session IDs, batch processing capabilities, and step-by-step research execution control.

🏗️ Architecture

The system follows a streamlined orchestrator-worker pattern with parallel execution:

                [User Query + Configuration]
                            │
                            ▼
           ┌─────────────────────────────────┐
           │       AdvancedResearch          │ (Main Orchestrator)
           │  - Session Management          │
           │  - Conversation History        │
           │  - Export Control              │
           └─────────────────────────────────┘
                            │ 1. Initialize Research Session
                            ▼
           ┌─────────────────────────────────┐
           │      Director Agent             │ (Research Coordinator)
           │  - Query Analysis & Planning    │
           │  - Task Decomposition           │
           │  - Research Strategy            │
           └─────────────────────────────────┘
                            │ 2. Decompose into Sub-Tasks
                            ▼
       ┌─────────────────────────────────────────┐
       │     Parallel Worker Execution           │
       │   (ThreadPoolExecutor - Concurrent)     │
       └─────────────────────────────────────────┘
          │           │           │           │
          ▼           ▼           ▼           ▼
    ┌──────────┐ ┌──────────┐ ┌──────────┐ ┌──────────┐
    │Worker 1  │ │Worker 2  │ │Worker 3  │ │Worker N  │
    │Exa Search│ │Exa Search│ │Exa Search│ │Exa Search│
    │Integration│ │Integration│ │Integration│ │Integration│
    └──────────┘ └──────────┘ └──────────┘ └──────────┘
          │           │           │           │
          ▼           ▼           ▼           ▼
       ┌─────────────────────────────────────────┐
       │      Results Aggregation                │
       │  - Combine Worker Outputs               │
       │  - Format Research Findings             │
       └─────────────────────────────────────────┘
                            │ 3. Synthesize Results
                            ▼
           ┌─────────────────────────────────┐
           │    Conversation Management      │
           │  - History Tracking             │
           │  - Output Formatting            │
           │  - Export Processing            │
           └─────────────────────────────────┘
                            │ 4. Deliver Results
                            ▼
              [Formatted Report + Optional JSON Export]

🔄 Workflow Process

1. Session Initialization: AdvancedResearch creates a unique research session with conversation tracking
2. Director Agent Planning: The director agent analyzes the query and plans research strategy
3. Parallel Worker Execution: Multiple worker agents execute concurrent searches using Exa API
4. Results Aggregation: Worker outputs are combined and synthesized into comprehensive findings
5. Output Processing: Results are formatted according to specified output type (JSON, markdown, etc.)
6. Export & Delivery: Optional JSON export with timestamped files and conversation history

🤝 Contributing

This implementation is part of the open-source swarms ecosystem. We welcome contributions!

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-research-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-research-feature)
5. Open a Pull Request

Development Setup with uv

# Clone and setup development environment
git clone https://github.com/The-Swarm-Corporation/AdvancedResearch.git
cd AdvancedResearch

uv venv

uv pip install -r requirements.txt

📄 License

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

📚 Citation

If you use this work in your research, please cite both the original paper and this implementation:

@misc{anthropic2024researchsystem,
    title={How we built our multi-agent research system},
    author={Anthropic},
    year={2024},
    month={June},
    url={https://www.anthropic.com/engineering/built-multi-agent-research-system}
}

@software{advancedresearch2024,
    title={AdvancedResearch: Enhanced Multi-Agent Research System},
    author={The Swarm Corporation},
    year={2024},
    url={https://github.com/The-Swarm-Corporation/AdvancedResearch},
    note={Implementation based on Anthropic's multi-agent research system paper}
}

@software{swarms_framework,
    title={Swarms: An Open-Source Multi-Agent Framework},
    author={Kye Gomez},
    year={2023},
    url={https://github.com/kyegomez/swarms}
}

📚 Documentation

For comprehensive API documentation, examples, and advanced usage:

📖 View Full API Documentation →

🔗 Related Work

• Original Paper - "How we built our multi-agent research system" by Anthropic
• Swarms Framework - The underlying multi-agent AI orchestration framework

📞 Support

• Issues: GitHub Issues
• Discussions: GitHub Discussions
• Discord: Join our community

Built with Swarms framework for production-grade agentic applications