Agentcy: Multi-Agent Collaboration with AutoGen

This code demonstrates the power of multi-agent collaboration using the AutoGen library. Instead of relying on a single agent to handle tasks, multiple specialized agents work together, each bringing its expertise to the table.

📖 Overview

The code sets up a collaborative environment where multiple agents, each with its unique role and expertise, come together to discuss, plan, and execute tasks. This collaboration ensures that different aspects of a task are handled by the most qualified agent, leading to more efficient and accurate outcomes.

🕵🏽 Agents

Here are the agents involved in the collaboration:

Client: Represents the user. Interacts with the Account Manager to discuss and approve the plan.
Account Manager: Nurtures the relationship between the agency and the client and ensures clear communication.
Strategist: Drafts strategic briefs that effectively position the client's brand in the market based on comprehensive research and insights.
Researcher: Delves deep into understanding user pain points, identifies market opportunities, and analyzes prevailing market conditions.
Marketer: Transforms strategy and insights into compelling marketable ideas that resonate with the target audience.
Manager: Oversees the entire project lifecycle, ensuring all agents are effectively fulfilling their objectives and tasks on time.
Designer: Transforms strategic and marketing ideas into compelling visual narratives.
Copywriter: Crafts compelling narratives and messages that align with the brand's strategy.
Media Planner: Identifies the best mix of media channels to deliver advertising messages.
Director: Guides the creative vision of the project, ensuring ideas are unique, compelling, and meet the highest standards of excellence.

🤝 Collaboration Flow

The GroupChat class is used to create a collaborative environment where all agents can communicate.
The GroupChatManager manages the group chat, ensuring smooth communication between agents.
The initiate_chat method is called to start the collaboration.

⚙️ Setup & Configuration

Ensure you have the required libraries installed:

pip install pyautogen

Set up the OpenAI configuration list by either providing an environment variable OAI_CONFIG_LIST or specifying a file path.

[
    {
        "model": "gpt-3.5-turbo", #or whatever model you prefer
        "api_key": "INSERT_HERE"
    }
]

Instantiate each agent with its unique name, configuration, and system message.
Create a group chat with all the agents and initiate the collaboration.

⏯️ Conclusion