🦜 AI-Parrot

A unified Python library for building intelligent agents, chatbots, and LLM-powered applications

AI-Parrot simplifies working with Large Language Models by providing a cohesive framework for creating conversational agents, managing tools, implementing RAG systems, and orchestrating complex AI workflows—all without the bloat of traditional frameworks.

✨ Key Features

🤖 Multi-Provider LLM Support

Connect seamlessly to multiple AI providers through a unified interface:

• OpenAI (GPT-4, GPT-3.5)
• Anthropic Claude (Claude 3.5 Sonnet, Opus)
• Google GenAI (Gemini models)
• Groq (Fast inference)

🛠️ Intelligent Agent System

Build sophisticated agents with built-in tool support and orchestration:

• Tool Manager: Share tools across multiple agents
• Agent Registry: Decorator-based agent creation and registration
• Python Tool Calling: Native support for calling Python functions as tools
• Complex Toolkits: Compose multiple tools into reusable toolkits

💬 Chatbot Creation

Create production-ready chatbots with minimal code:

• Conversational context management
• Multi-turn dialogue support
• Streaming responses
• Custom personality and behavior configuration

🗄️ Knowledge Base & RAG

Implement Retrieval-Augmented Generation with enterprise-grade components:

• PgVector Integration: PostgreSQL-based vector storage for semantic search
• Document Loaders: Transform any document format into AI-ready context
• Open-Source Embeddings: Hugging Face Transformers integration
• Structured Outputs: Type-safe responses from your LLMs

🌐 API & Server Capabilities

Deploy your AI applications with ease:

• Bot Manager: Centralized management for multiple bot instances
• REST API: Expose your agents and chatbots via HTTP endpoints
• MCP Server: Model Context Protocol support for standardized agent communication

⏰ Task Scheduling

Orchestrate agent actions over time:

• Schedule periodic agent tasks
• Trigger-based automation
• Asynchronous execution support
• Task dependency management

🚀 Quick Start

Installation

pip install ai-parrot

Create Your First Chatbot

from ai_parrot import ChatBot, OpenAIClient

# Initialize LLM client
client = OpenAIClient(api_key="your-api-key")

# Create a chatbot
bot = ChatBot(
    name="assistant",
    client=client,
    system_prompt="You are a helpful AI assistant."
)

# Have a conversation
response = bot.chat("What's the weather like today?")
print(response)

Build an Agent with Tools

from ai_parrot import Agent, tool
from ai_parrot.registry import agent_registry

@tool
def calculate_sum(a: int, b: int) -> int:
    """Add two numbers together."""
    return a + b

@tool
def get_current_time() -> str:
    """Get the current time."""
    from datetime import datetime
    return datetime.now().strftime("%H:%M:%S")

# Register an agent with tools
@agent_registry.register("math_agent")
class MathAgent(Agent):
    def __init__(self):
        super().__init__(
            name="Math Helper",
            tools=[calculate_sum, get_current_time]
        )

# Use the agent
agent = agent_registry.get("math_agent")
result = agent.run("What's 42 plus 58? Also, what time is it?")

Implement RAG with Vector Store

from ai_parrot import RAGChatBot, PgVectorStore
from ai_parrot.loaders import PDFLoader, TextLoader

# Initialize vector store
vector_store = PgVectorStore(
    connection_string="postgresql://user:pass@localhost/db"
)

# Load and index documents
loader = PDFLoader()
documents = loader.load("./docs/manual.pdf")
vector_store.add_documents(documents)

# Create RAG-enabled chatbot
rag_bot = RAGChatBot(
    client=client,
    vector_store=vector_store,
    top_k=5
)

# Query with context
response = rag_bot.chat("How do I configure the settings?")

Expose via API

from ai_parrot import BotManager, create_api

# Create bot manager
manager = BotManager()
manager.register_bot("assistant", bot)
manager.register_agent("math_helper", agent)

# Create and run API server
app = create_api(manager)

# Run with: uvicorn main:app --reload

Schedule Agent Tasks

from ai_parrot import TaskScheduler

scheduler = TaskScheduler()

# Schedule a daily summary
@scheduler.schedule(cron="0 9 * * *")  # Every day at 9 AM
async def daily_summary():
    summary = await agent.run("Generate a summary of yesterday's activities")
    send_email(summary)

# Run the scheduler
scheduler.start()

🏗️ Architecture

AI-Parrot is designed with modularity and extensibility in mind:

┌─────────────────────────────────────────┐
│          Application Layer              │
│  (Chatbots, Agents, Custom Logic)       │
└──────────────┬──────────────────────────┘
               │
┌──────────────▼──────────────────────────┐
│         AI-Parrot Core                  │
│  • Agent Registry  • Tool Manager       │
│  • Bot Manager    • Task Scheduler      │
└──────────────┬──────────────────────────┘
               │
┌──────────────▼──────────────────────────┐
│      Provider Integrations              │
│  • OpenAI    • Claude    • Gemini       │
│  • Groq      • Hugging Face             │
└──────────────┬──────────────────────────┘
               │
┌──────────────▼──────────────────────────┐
│      Storage & Infrastructure           │
│  • PgVector  • Document Loaders         │
│  • MCP Server • API Layer               │
└─────────────────────────────────────────┘

🎯 Use Cases

• Customer Support Bots: Build intelligent support agents with knowledge base integration
• Research Assistants: Create agents that can search, analyze, and synthesize information
• Automation Workflows: Schedule and orchestrate AI-powered tasks
• Internal Tools: Expose LLM capabilities through APIs for your team
• Multi-Agent Systems: Coordinate multiple specialized agents working together

🗺️ Roadmap

• ✅ Langchain Independence: Removed heavyweight dependencies
• 🚧 Complex Toolkits: Advanced tool composition and chaining
• 🚧 Model Interoperability: Seamless LLM + Hugging Face model integration
• 📋 Non-LLM Models: Support for classification, NER, and other ML models
• 📋 MCP Full Integration: Complete Model Context Protocol implementation
• 📋 Graph-Based RAG: Knowledge graphs with ArangoDB for advanced reasoning

🤝 Contributing

Contributions are welcome! Whether it's bug fixes, new features, or documentation improvements, we appreciate your help in making AI-Parrot better.

📄 License

MIT License.

📚 Documentation

For detailed documentation, examples, and API reference, see the examples/ folder.

💬 Community & Support

• Issues: GitHub Issues
• Discussions: GitHub Discussions

---

Built with ❤️ for developers who want powerful AI tools without the complexity.