promptopt 0.1.0

Enterprise prompt optimization framework combining DSPy and GRPO approaches

PromptOpt: Enterprise Prompt Optimization Framework

A unified framework for testing, comparing, and hybridizing DSPy and GRPO approaches to prompt optimization, with special focus on enterprise deployment via synthetic data and Colab accessibility.

🌟 Key Features

• 🤖 Multiple Optimization Strategies: DSPy, GRPO, and hybrid approaches
• 💰 Cost-Aware Optimization: Built-in budget management and tracking
• 🏢 Enterprise-Ready: Compliance support, ROI analysis, deployment tools
• ☁️ Colab-Optimized: Works perfectly in Google Colab (no GPU needed)
• 📊 Tournament Evaluation: Head-to-head prompt comparison system
• 🎯 Synthetic Data: Generate realistic business scenarios
• 🔧 API-Based: Works with OpenAI and Anthropic APIs

Installation

Using Conda (Recommended)

# Create a new conda environment
conda create -n promptopt python=3.9
conda activate promptopt

# Install the package
pip install -e .

# For development
pip install -r requirements-dev.txt

Using pip only

pip install promptopt

For development:

pip install -e ".[dev]"

For Colab environments:

pip install promptopt[colab]

Quick Start

from promptopt import EnterprisePOC
from promptopt.colab import ColabManager

# Set up environment
manager = ColabManager()
manager.setup_enterprise_environment()

# Run a complete POC
poc = EnterprisePOC()
results = poc.run_complete_poc(
    business_scenario="customer_support",
    company_context={"industry": "tech", "size": "enterprise"},
    budget_limit=500.0
)

# View results
print(f"Optimization improvements: {results.optimization_improvements}")
print(f"Projected ROI: {results.roi_projections}")

📚 Documentation

See DOCUMENTATION.md for comprehensive guides and API reference.

🚀 Quick Start in Google Colab

🧪 Running Tests

conda activate promptopt
pytest tests/

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

📄 License

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

🙏 Acknowledgments

• DSPy framework for few-shot optimization techniques
• GRPO methodology for tournament-based optimization
• OpenAI and Anthropic for LLM APIs

📊 Example Results

Using PromptOpt, teams typically see:

• 30-40% improvement in response quality
• 85%+ consistency across team members
• $2000+/month savings from optimized prompts
• 2-week ROI for enterprise deployments