grami-ai 0.2.0

Async AI Agent Framework

🤖 Grami AI: Async AI Agent Framework

🌟 Overview

Grami AI is a cutting-edge, flexible Python framework for building intelligent, asynchronous AI agents with comprehensive Language Model (LLM) integration.

🚀 Key Features

• 🔄 Fully Asynchronous Design
• 🧩 Modular Architecture
• 🔒 Thread-Safe Operations
• 📦 Multiple Backend Support
• 🛠 Extensible Tool Ecosystem

📦 Installation

pip install grami-ai

🛠 Tools Ecosystem

Base Tools

Grami AI provides a robust set of async tools for various operations:

Calculator Tool

from grami_ai.tools import CalculatorTool

async def calculate():
    calculator = CalculatorTool()
    result = await calculator.execute('2 + 3 * 4')  # Returns 14.0

JSON Parser Tool

from grami_ai.tools import JSONParserTool

async def parse_json():
    json_tool = JSONParserTool()
    parsed = await json_tool.execute('{"name": "John"}')
    filtered = await json_tool.execute(
        '{"name": "John", "age": 30}', 
        operation='transform', 
        filter_keys=['name']
    )

String Manipulation Tool

from grami_ai.tools import StringManipulationTool

async def manipulate_string():
    string_tool = StringManipulationTool()
    
    # Clean text
    cleaned = await string_tool.execute("  Hello   World!  ")
    
    # Count words
    word_count = await string_tool.execute("Text", operation='count_words')
    
    # Reverse text
    reversed_text = await string_tool.execute("Hello", operation='reverse')
    
    # Capitalize text
    capitalized = await string_tool.execute(
        "hello world", 
        operation='capitalize', 
        mode='all'
    )

Web Scraper Tool

from grami_ai.tools import WebScraperTool

async def scrape_web():
    web_tool = WebScraperTool()
    
    # Fetch content
    content = await web_tool.execute('https://example.com')
    
    # Parse text
    parsed_text = await web_tool.execute(
        'https://example.com', 
        operation='parse'
    )

Memory Providers

In-Memory Memory

from grami_ai.memory import AsyncInMemoryMemory

async def use_memory():
    memory = AsyncInMemoryMemory(max_size=100)
    await memory.add_item('conversation_1', {'role': 'user', 'content': 'Hello'})
    items = await memory.get_items('conversation_1')

Redis Memory

from grami_ai.memory import AsyncRedisMemory

async def use_redis_memory():
    memory = AsyncRedisMemory(redis_url='redis://localhost:6379')
    await memory.add_item('conversation_1', {'role': 'user', 'content': 'Hello'})
    items = await memory.get_items('conversation_1')

🚀 Comprehensive Examples

Grami AI provides rich, real-world example scripts to demonstrate the framework's capabilities:

1. 🔬 AI Research Assistant

examples/ai_research_assistant.py

A sophisticated research workflow that showcases:

• Async tool composition
• Web scraping
• Text processing
• Memory management
• Complex research automation

async def main():
    research_assistant = AIResearchAssistant()
    research_results = await research_assistant.research_topic(
        "Artificial Intelligence in Scientific Research"
    )

2. 📊 Data Analysis Pipeline

examples/data_analysis_pipeline.py

An advanced data analysis tool demonstrating:

• Stock data retrieval
• Web scraping
• JSON parsing
• Statistical calculations
• Async data processing

async def main():
    pipeline = DataAnalysisPipeline()
    analysis_results = await pipeline.analyze_stock_data('AAPL')

3. 🌐 Multi-Tool Global Trend Analysis

examples/multi_tool_workflow.py

A complex workflow showcasing:

• Parallel topic research
• Async data gathering
• Comprehensive trend analysis
• Tool interaction and synthesis

async def main():
    workflow = MultiToolWorkflow()
    global_trends = await workflow.analyze_global_trends([
        "Artificial Intelligence",
        "Climate Change",
        "Renewable Energy"
    ])

🖥️ Running Examples

To run the example scripts, ensure you have the necessary dependencies:

# Install optional dependencies for web scraping
pip install aiohttp beautifulsoup4

# Run specific example scripts
python examples/ai_research_assistant.py
python examples/data_analysis_pipeline.py
python examples/multi_tool_workflow.py

Note: Some examples involve web scraping and may require:

• Active internet connection
• Handling potential network-related exceptions
• Respecting website terms of service

🎓 Learning Paths

These examples demonstrate different aspects of the Grami AI framework:

• Beginners: Start with individual tool usage
• Intermediate: Explore tool composition
• Advanced: Study complex async workflows

🚀 Quick Start

1. Clone the repository
2. Install dependencies
3. Explore and run example scripts
4. Modify and adapt to your use cases

git clone https://github.com/grami-ai/framework.git
cd grami-ai
pip install -e .
python examples/ai_research_assistant.py

💡 Example Highlights

• Async-First Design: Non-blocking, high-performance operations
• Modular Architecture: Easily extensible and customizable
• Tool Composition: Seamless integration of multiple tools
• Memory Management: Efficient context tracking
• Error Handling: Robust and informative error management

📋 Requirements

• Python 3.8+
• asyncio
• aioredis (optional)
• aiohttp (optional)
• beautifulsoup4 (optional)

🔧 Development

Setup

git clone https://github.com/grami-ai/framework.git
cd grami-ai
pip install -e .[dev]

Running Tests

pytest

🤝 Contributing

We welcome contributions! Please see our contribution guidelines.

📄 License

MIT License

🌐 Project Links

• Documentation: GitHub README
• Source Code: GitHub Repository
• Issue Tracker: GitHub Issues

💡 Future Roadmap

• Expand LLM provider implementations
• Enhance tool ecosystem
• Develop more memory backends
• Improve documentation
• Add comprehensive testing
• Performance optimization
• More example use cases