Invicta AI allows you to assemble individual agents into complex teams, leveraging their unique capabilities to collaborate on intricate tasks. Here’s how you can effectively create and manage AI Teammates.

​
Overview

AI Teammates enable agents to invoke other agents in a manner similar to tool invocation. This facilitates the sharing of information and allows agents to command each other or use tools available to their teammates. Agents rely on the names and descriptions of their teammates to decide when to trigger them during interactions.
Instruct an agent to invoke its AI teammate by simply providing the teammate’s name, or allow the agent to deduce the need based on the teammate’s description.

​
How It Works

  1. When an agent’s system message or your instructions trigger it, it sends a request with relevant information to another agent.
  2. The invoked agent uses its “Preferred LLM” to generate an appropriate response.
You can add AI Teammates to any given agent in the Agent Builder.
You can only select agents that belong to you as AI teammates.

​
Visual Representation

If an agent is a teammate of other agents (upstream) or has other teammates (downstream), they will be displayed visually.
For cloning agent teams, refer to Cloning Agent Teams.

​
Conversation Patterns

You can design various conversation patterns using AI Teammates:
  • Hierarchical: Tasks are delegated from a primary agent down a chain of command.
  • Sequential: Tasks are passed from one agent to another in a specific order.
  • Cyclical: Tasks circulate among agents in a loop until completion.
There is a maximum iteration limit of 15. If one agent invokes an agent 15 times or 15 different agents once, it will be stopped automatically to prevent loops.

​
Benefits of AI Agent Teams

Using AI Teammates offers several key advantages:
  • Specialization: Each agent can focus on specific tasks, enhancing efficiency and output quality.
  • Cost-Effectiveness and Scalability: Smaller, task-specific agents reduce errors and operational costs while scaling efficiently.
  • Enhanced Accuracy through Collaboration: Distributed tasks among specialized agents improve data processing and decision-making accuracy.
  • Segmented Knowledge Management: Isolated data handling enhances security and ensures data integrity.
  • Modularity: Simplifies updates and maintenance, facilitating agile development and reuse across projects.
  • Collaborative Learning: Agents continuously refine their problem-solving approaches through inter-agent feedback, leveraging collective intelligence for innovation.
For more detailed information on the benefits of AI Agent Teams, please read the full article here.