augini 0.3.1

AI-powered Python framework for tabular data enrichment and analysis using LLMs. Features include intelligent feature engineering, natural language data analysis, and AI agents for automated workflows.

Augini 🤖

🎯 What is Augini?

Augini is an AI-powered Python framework for tabular data enrichment and analysis. It leverages Large Language Models (LLMs) to:

• Generate meaningful features from your data
• Provide natural language data analysis
• Create AI agents for automated data workflows

🚀 Quick Start

pip install augini

from augini import DataEngineer, DataAnalyzer
import pandas as pd

# Sample customer data
df = pd.DataFrame({
    'CustomerID': ['C001', 'C002'],
    'Age': [25, 45],
    'MonthlyCharges': [50.0, 75.0]
})

# Initialize with your API key (supports OpenAI, OpenRouter, Azure)
engineer = DataEngineer(
    api_key="your-api-key",
    model="gpt-4o-mini",  # Use OpenRouter's GPT-4
    base_url="https://openrouter.ai/api/v1"  # Optional: use OpenRouter
)

# Generate customer insights
df = engineer.generate_features(
    df=df,
    features=[
        {
            'new_feature_name': 'CustomerSegment',
            'new_feature_description': 'Classify customer segment based on age and spending',
            'output_type': 'category',
            'constraints': {'categories': ['Premium', 'Regular', 'Budget']}
        },
        {
            'new_feature_name': 'ChurnRisk',
            'new_feature_description': 'Calculate churn risk score (0-100)',
            'output_type': 'float',
            'constraints': {'min': 0, 'max': 100}
        }
    ]
)

# Initialize analyzer for natural language insights
analyzer = DataAnalyzer(
    api_key="your-api-key",
    model="gpt-4o-mini",
    enable_memory=True  # Enable conversation context
)

# Fit data and ask questions
analyzer.fit(df)
insights = analyzer.chat("What patterns do you see in customer segments?")
print(insights)

🎁 Key Features

🔄 DataEngineer

• Feature Generation: Create meaningful features using AI
• Data Augmentation: Enrich datasets with synthetic data
• Custom Constraints: Control output formats and ranges
• Batch Processing: Handle large datasets efficiently

📊 DataAnalyzer

• Natural Language Analysis: Ask questions about your data
• Pattern Detection: Uncover hidden trends and correlations
• Memory Context: Build on previous analysis
• Visualization Integration: Generate plots and charts

🤖 AI Agents

• Automated Workflows: Create agents for repetitive tasks
• Custom Behaviors: Define agent goals and constraints
• Chain Actions: Connect multiple agents for complex workflows

🌐 Provider Agnostic

Augini works with multiple LLM providers:

• OpenAI
• OpenRouter
• Azure OpenAI
• Anthropic (coming soon)

🤝 Contributing

We welcome contributions!

📜 License

Augini is released under the MIT License.