GRAMI-AI: Dynamic AI Agent Framework

📋 Table of Contents

• Overview
• Key Features
• Installation
• Quick Start
• Provider Examples
• Memory Management
• Streaming Capabilities
• Development Roadmap
• TODO List
• Contributing
• License

🌟 Overview

GRAMI-AI is a cutting-edge, async-first AI agent framework designed for building sophisticated AI applications. With support for multiple LLM providers, advanced memory management, and streaming capabilities, GRAMI-AI enables developers to create powerful, context-aware AI systems.

Why GRAMI-AI?

• Async-First: Built for high-performance asynchronous operations
• Provider Agnostic: Support for Gemini, OpenAI, Anthropic, and Ollama
• Advanced Memory: LRU and Redis-based memory management
• Streaming Support: Efficient token-by-token streaming responses
• Enterprise Ready: Production-grade security and scalability

🚀 Key Features

LLM Providers

• Gemini (Google's latest LLM)
• OpenAI (GPT models)
• Anthropic (Claude)
• Ollama (Local models)

Memory Management

• LRU Memory (In-memory caching)
• Redis Memory (Distributed caching)
• Custom memory providers

Communication

• Synchronous messaging
• Asynchronous streaming
• WebSocket support
• Custom interfaces

💻 Installation

pip install grami-ai

🔑 API Key Setup

Before using GRAMI-AI, you need to set up your API keys. You can do this by setting environment variables:

export GEMINI_API_KEY="your-gemini-api-key"
# Or for other providers:
export OPENAI_API_KEY="your-openai-api-key"
export ANTHROPIC_API_KEY="your-anthropic-api-key"

Or using a .env file:

GEMINI_API_KEY=your-gemini-api-key
OPENAI_API_KEY=your-openai-api-key
ANTHROPIC_API_KEY=your-anthropic-api-key

🎯 Quick Start

Here's a simple example of how to create an AI agent using GRAMI-AI:

from grami.agents import AsyncAgent
from grami.providers.gemini_provider import GeminiProvider
from grami.memory.lru import LRUMemory
import asyncio
import os

async def main():
    # Initialize memory and provider
    memory = LRUMemory(capacity=5)
    provider = GeminiProvider(
        api_key=os.getenv("GEMINI_API_KEY"),
        generation_config={
            "temperature": 0.9,
            "top_p": 0.9,
            "top_k": 40,
            "max_output_tokens": 1000,
            "candidate_count": 1
        }
    )
    
    # Create agent
    agent = AsyncAgent(
        name="MyAssistant",
        llm=provider,
        memory=memory,
        system_instructions="You are a helpful AI assistant."
    )
    
    # Example: Using streaming responses
    message = "Tell me a short story about AI."
    async for chunk in agent.stream_message(message):
        print(chunk, end="", flush=True)
    print("\n")
    
    # Example: Using non-streaming responses
    response = await agent.send_message("What's the weather like today?")
    print(f"Response: {response}")

if __name__ == "__main__":
    asyncio.run(main())

📚 Provider Examples

Gemini Provider

from grami.providers.gemini_provider import GeminiProvider
from grami.memory.lru import LRUMemory

# Initialize with memory
provider = GeminiProvider(
    api_key="YOUR_API_KEY",
    model="gemini-pro",  # Optional, defaults to gemini-pro
    generation_config={   # Optional
        "temperature": 0.7,
        "top_p": 0.8,
        "top_k": 40
    }
)

# Add memory provider
memory = LRUMemory(capacity=100)
provider.set_memory_provider(memory)

# Regular message
response = await provider.send_message("What is AI?")

# Streaming response
async for chunk in provider.stream_message("Tell me a story"):
    print(chunk, end="", flush=True)

🧠 Memory Management

LRU Memory

from grami.memory.lru import LRUMemory

# Initialize with capacity
memory = LRUMemory(capacity=100)

# Add to agent
agent = AsyncAgent(
    name="MemoryAgent",
    llm=provider,
    memory=memory
)

Redis Memory

from grami.memory.redis import RedisMemory

# Initialize Redis memory
memory = RedisMemory(
    host="localhost",
    port=6379,
    capacity=1000
)

# Add to provider
provider.set_memory_provider(memory)

🌊 Streaming Capabilities

Basic Streaming

async def stream_example():
    async for chunk in provider.stream_message("Generate a story"):
        print(chunk, end="", flush=True)

Streaming with Memory

async def stream_with_memory():
    # First message
    response = await provider.send_message("My name is Alice")
    
    # Stream follow-up (will remember context)
    async for chunk in provider.stream_message("What's my name?"):
        print(chunk, end="", flush=True)

🗺 Development Roadmap

Core Framework Design

• Implement AsyncAgent base class with dynamic configuration
• Create flexible system instruction definition mechanism
• Design abstract LLM provider interface
• Develop dynamic role and persona assignment system
• Comprehensive async example configurations
  • Memory with streaming
  • Memory without streaming
  • No memory with streaming
  • No memory without streaming
• Implement multi-modal agent capabilities (text, image, video)

LLM Provider Abstraction

• Unified interface for diverse LLM providers
• Google Gemini integration
  • Basic message sending
  • Streaming support
  • Memory integration
• OpenAI ChatGPT integration
  • Basic message sending
  • Streaming implementation
  • Memory support
• Anthropic Claude integration
• Ollama local LLM support
• Standardize function/tool calling across providers
• Dynamic prompt engineering support
• Provider-specific configuration handling

Communication Interfaces

• WebSocket real-time communication
• REST API endpoint design
• Kafka inter-agent communication
• gRPC support
• Event-driven agent notification system
• Secure communication protocols

Memory and State Management

• Pluggable memory providers
• In-memory state storage (LRU)
• Redis distributed memory
• DynamoDB scalable storage
• S3 content storage
• Conversation and task history tracking
• Global state management for agent crews
• Persistent task and interaction logs
• Advanced memory indexing
• Memory compression techniques

Tool and Function Ecosystem

• Extensible tool integration framework
• Default utility tools
  • Kafka message publisher
  • Web search utility
  • Content analysis tool
• Provider-specific function calling support
• Community tool marketplace
• Easy custom tool development

Agent Crew Collaboration

• Inter-agent communication protocol
• Workflow and task delegation mechanisms
• Approval and review workflows
• Notification and escalation systems
• Dynamic team composition
• Shared context and memory management

Use Case Implementations

• Digital Agency workflow template
  • Growth Manager agent
  • Content Creator agent
  • Trend Researcher agent
  • Media Creation agent
• Customer interaction management
• Approval and revision cycles

Security and Compliance

• Secure credential management
• Role-based access control
• Audit logging
• Compliance with data protection regulations

Performance and Scalability

• Async-first design
• Horizontal scaling support
• Performance benchmarking
• Resource optimization

Testing and Quality

• Comprehensive unit testing
• Integration testing for agent interactions
• Mocking frameworks for LLM providers
• Continuous integration setup

Documentation and Community

• Detailed API documentation
• Comprehensive developer guides
• Example use case implementations
• Contribution guidelines
• Community tool submission process
• Regular maintenance and updates

Future Roadmap

• Payment integration solutions
• Advanced agent collaboration patterns
• Specialized industry-specific agents
• Enhanced security features
• Extended provider support

📝 TODO List

• Add support for Gemini provider
• Implement advanced caching strategies (LRU)
• Add WebSocket support for real-time communication
• Create comprehensive test suite
• Add support for function calling
• Implement conversation branching
• Add support for multi-modal inputs
• Enhance error handling and logging
• Add rate limiting and quota management
• Create detailed API documentation
• Add support for custom prompt templates
• Implement conversation summarization
• Add support for multiple languages
• Implement fine-tuning capabilities
• Add support for model quantization
• Create a web-based demo
• Add support for batch processing
• Implement conversation history export/import
• Add support for custom model hosting
• Create visualization tools for conversation flows
• Implement automated testing pipeline
• Add support for conversation analytics
• Create deployment guides for various platforms
• Implement automated documentation generation
• Add support for model performance monitoring
• Create benchmarking tools
• Implement A/B testing capabilities
• Add support for custom tokenizers
• Create model evaluation tools
• Implement conversation templates
• Add support for conversation routing
• Create debugging tools
• Implement conversation validation
• Add support for custom memory backends
• Create conversation backup/restore features
• Implement conversation filtering
• Add support for conversation tagging
• Create conversation search capabilities
• Implement conversation versioning
• Add support for conversation merging
• Create conversation export formats
• Implement conversation import validation
• Add support for conversation scheduling
• Create conversation monitoring tools
• Implement conversation archiving
• Add support for conversation encryption
• Create conversation access control
• Implement conversation rate limiting
• Add support for conversation quotas
• Create conversation usage analytics
• Implement conversation cost tracking
• Add support for conversation billing
• Create conversation audit logs
• Implement conversation compliance checks
• Add support for conversation retention policies
• Create conversation backup strategies
• Implement conversation recovery procedures
• Add support for conversation migration
• Create conversation optimization tools
• Implement conversation caching strategies
• Add support for conversation compression
• Create conversation performance metrics
• Implement conversation health checks
• Add support for conversation monitoring
• Create conversation alerting system
• Implement conversation debugging tools
• Add support for conversation profiling
• Create conversation testing framework
• Implement conversation documentation
• Add support for conversation examples
• Create conversation tutorials
• Implement conversation guides
• Add support for conversation best practices
• Create conversation security guidelines

🤝 Contributing

We welcome contributions! Please feel free to submit a Pull Request.

📄 License

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

🔗 Links

• PyPI Package
• GitHub Repository
• Documentation

📧 Support

For support, email support@yafatek.dev or create an issue on GitHub.