Praison AI: Ultimate Multi-Agent System for LLM Apps

Praison AI emerges as a groundbreaking low-code, centralized framework that leverages AutoGen, CrewAI, and other agent frameworks to simplify the creation and orchestration of multi-agent systems. This innovative solution is designed for various Large Language Model (LLM) applications, emphasizing ease of use, customization, and human-AI interaction.

Key Features and Capabilities

Praison AI boasts an impressive array of features:

• Automated AI Agent Creation: Streamlines the process of generating AI agents
• Flexible Framework Support: Utilizes CrewAI, AutoGen, or other agent frameworks
• Extensive LLM Compatibility: Supports over 100 Large Language Models
• Code Repository Interaction: Enables chatting with entire codebases
• Interactive User Interface: Provides a user-friendly graphical interface
• YAML-Based Configuration: Simplifies setup and customization
• Custom Tool Integration: Allows for the incorporation of specialized tools

User Interfaces

Praison AI offers multiple user interfaces to cater to different needs:

1. Multi-Agent UI: For CrewAI or AutoGen (https://docs.praison.ai/ui/ui)
2. Chat UI: For conversing with 100+ LLMs using a single AI agent (https://docs.praison.ai/ui/chat)
3. Code UI: For interacting with entire codebases using a single AI agent (https://docs.praison.ai/ui/code)

Praison AI builds upon the work done by CrewAI, which itself is based on Langchain. As a result, Praison AI emerges as an even more low-code, centralized framework. Its primary goal is to simplify the creation and orchestration of multi-agent systems for various LLM applications, with a strong emphasis on ease of use, customization, and human-AI interaction.

The framework provides examples of using both AutoGen and CrewAI, allowing users to switch between these two powerful frameworks as needed. This flexibility enables developers to leverage the strengths of each framework depending on their specific use case.

User Interface and Operation Modes

Praison AI offers a graphical user interface that automates agent operations. Users can simply input their desired tasks into a dialogue box, prompting automatic execution. This feature significantly reduces the learning curve and allows even non-technical users to harness the power of multi-agent AI systems.

The system provides two primary operating modes:

Automatic Mode: Executes agent operations automatically based on user input

Manual Mode: Allows for customization of agents and tools

In manual mode, users have the ability to fine-tune agents and integrate custom tools, providing a high degree of flexibility for advanced users and complex projects.

Integrated Tools and Model Support

Praison AI integrates several powerful tools to enhance its capabilities:

• mem0: A sophisticated memory layer for improved context retention
• Craw4AI and Firecrawl: Two AI-powered web crawling tools for efficient data gathering
• Langchain: A well-known framework for developing applications powered by language models

The framework also offers support for a wide range of AI models, ensuring compatibility with various LLM ecosystems. This extensive support allows users to leverage the strengths of different models for specific tasks.

Installation and Setup

Basic Installation

pip install praisonai

Specialized Installations

• For code-related features: pip install "praisonai[code]"
• For chat features: pip install "praisonai[chat]"
• For UI features: pip install -U "praisonai[ui]"

API Key Setup

export OPENAI_API_KEY="Enter your API key"

Note: You can generate your OpenAI API key at https://platform.openai.com/api-keys. Alternative providers like Ollama and Mistral are also supported.

Quick Start Guide

1. Initialization:

   praisonai --init create a movie script about dog in moon

This creates an agents.yaml file in the current directory.

2. Running:

   praisonai

or

   python -m praisonai

3. Specifying Agent Framework (optional):

   praisonai --framework autogen

4. Automatic Mode:

   praisonai --auto create a movie script about Dog in Moon

Advanced Features

User Interface Setup

For the Chainlit-based UI:

export OPENAI_API_KEY="Enter your API key"
chainlit create-secret
export CHAINLIT_AUTH_SECRET=xxxxxxxx
praisonai ui

Custom Tool Creation

Praison AI allows for the creation of custom tools, extending the framework’s capabilities. More information can be found at https://docs.praison.ai/tools/custom/.

Agent Scripting

Users can define agent behaviors using YAML-based scripts. Here’s a simple example:

framework: crewai
topic: Artificial Intelligence
roles:
  screenwriter:
    backstory: "Skilled in crafting scripts with engaging dialogue about {topic}."
    goal: Create scripts from concepts.
    role: Screenwriter
    tasks:
      scriptwriting_task:
        description: "Develop scripts with compelling characters and dialogue about {topic}."
        expected_output: "Complete script ready for production."

Integration in Projects

Praison AI can be easily integrated into existing projects. Here are two methods:

Option 1: Using Raw YAML

from praisonai import PraisonAI

agent_yaml = """
framework: "crewai"
topic: "Space Exploration"
roles:
  astronomer:
    role: "Space Researcher"
    goal: "Discover new insights about {topic}"
    backstory: "You are a curious and dedicated astronomer with a passion for unraveling the mysteries of the cosmos."
    tasks:
      investigate_exoplanets:
        description: "Research and compile information about exoplanets discovered in the last decade."
        expected_output: "A summarized report on exoplanet discoveries, including their size, potential habitability, and distance from Earth."
"""

praisonai = PraisonAI(agent_yaml=agent_yaml)
result = praisonai.run()
print(result)

Option 2: Using a Separate agents.yaml File

from praisonai import PraisonAI

def basic():
    praisonai = PraisonAI(agent_file="agents.yaml")
    praisonai.run()

if __name__ == "__main__":
    basic()

Development and Contribution

For developers interested in contributing or customizing Praison AI, the project uses Poetry for dependency management. Here are some useful commands:

• Install all dependencies: poetry install
• Install only documentation dependencies: poetry install --with docs
• Install only test dependencies: poetry install --with test
• Install only development dependencies: poetry install --with dev

Conclusion

Praison AI represents a significant advancement in multi-agent AI systems, offering a user-friendly, versatile solution for creating and managing complex AI agents. Its low-code approach, combined with extensive customization options and support for various LLMs, makes it an invaluable tool for developers, researchers, and businesses looking to harness the power of AI in their applications.