Core Concepts

Agents, Teams & Routers

FloAI is built around the idea of creating composable, flexible agentic workflows that interact with one another seamlessly. Each component in FloAI, from agents to teams, plays a specific role in building intelligent, reactive, and modular systems. The core philosophy revolves around making these components composable, allowing developers to easily extend, modify, and combine them into larger, more complex systems.

Basic Concepts

Agents

An Agent in FloAI is an autonomous entity that performs tasks by interacting with different environments, data sources, or APIs. Agents can have tools which can perform certain actions like sending an email or updating a database. FloAI supports different kinds of agents like llm agents, agentic agents, delegator agents, etc.

Tools

Tools are fundamental building blocks that enable AI agents to interact with the external world and perform specific tasks. Think of them as the "hands" of an agent - they allow the agent to take concrete actions rather than just process information.

Agent tools are specialized functions or interfaces that:

• Perform specific, well-defined actions

• Take structured inputs and return clear outputs

• Interface with external systems, APIs, or services

• Execute real-world tasks or data manipulations

Teams

Teams consist of a group of agents connected by a router. A team consists of multiple agents working together to accomplish a task. The router decides how the flow of control is within the team. Flo has different kinds of routers like supervisor, linear, or llm

Routers

Routers are the high-level component that controls the flow of execution within a team and even between teams. Depending on what is added to the router, routers can decide how agents/teams should be executed.

These three components are composable with each other to create bigger structures like teams of teams etc. Agents and teams are composable with each other through Routers