Platform documentation

Create and configure an agent in Fusioni Platform.

Learn how to create a new agent from the platform dashboard, choose its type, configure the LLM provider and model, connect tools and knowledge sources, and save it for workflows, chats, and automation pipelines.

Overview

Use this page when you need to create a new agent inside Fusioni Platform. The flow starts from the dashboard, continues through the Agents area, and ends when the agent configuration is saved.

Reusable agent setup

Create agents that can be reused across workflows, chats, and automation pipelines.

Controlled model behavior

Keep model choice, instructions, tools, and knowledge sources visible and configurable.

Context-aware responses

Give the agent access to approved knowledge so responses are more relevant and grounded.

Configuration fields

Before saving an agent, review the main configuration areas below. These fields define how the agent is identified, which model it uses, what actions it can perform, and what knowledge it can reference.

Agent identity

Give the agent a clear name and description so your team understands its purpose inside the workspace.

LLM configuration

Choose a provider and model that match the agent’s role, cost profile, and quality requirements.

Tools

Assign tools that let the agent call APIs, perform actions, or participate in platform workflows.

Knowledge sources

Connect storage or knowledge bases so the agent can answer with the right context.

Procedure

Create the agent

This guide follows the standard agent creation flow in Fusioni Platform. Start from the dashboard, open Agents, add a new agent, then configure identity, type, model, tools, knowledge, and instructions.

  1. 1

    Step 1

    Start from the dashboard

    Begin on the main dashboard, where you can monitor usage, token consumption, and activity across your organization.

  2. 2

    Step 2

    Open Agents from the side menu

    Use the left-hand navigation and select Agents to view the list of existing agents in your workspace.

  3. 3

    Step 3

    Review existing agents

    The Agents page shows each agent’s name, type, and the language model it uses, helping you understand what already exists before adding another one.

  4. 4

    Step 4

    Click Add Agent

    Select Add Agent to open the agent configuration screen and begin creating a new agent.

  5. 5

    Step 5

    Enter name, description, and type

    Add a required name, write a clear description of the agent’s purpose, and choose the agent type that matches your use case.

  6. 6

    Step 6

    Configure the LLM

    Choose the LLM provider and then select the specific model the agent should use. Fusioni supports providers such as OpenRouter, Together, Lambda, Fireworks, and more.

  7. 7

    Step 7

    Connect tools and knowledge sources

    Assign tools to extend the agent’s capabilities and connect storage or knowledge bases for context-aware answers.

  8. 8

    Step 8

    Add optional instructions

    Use additional instructions to guide the agent’s behavior, tone, boundaries, and response style for your workflow.

  9. 9

    Step 9

    Review and save

    Review the configuration and save the agent. It is then ready to use in workflows, chats, or automation pipelines.

Use the agent across the platform.

Once saved, your agent can become part of larger platform experiences. Use it alone in chat, connect it to journeys, or combine it with tools and knowledge bases for end-to-end automation.

Workflows

Add the agent to journeys where it can plan, retrieve context, call tools, or hand off to another step.

Chats

Use the agent in conversational experiences where users ask questions and receive context-aware replies.

Automation pipelines

Connect the agent to repeatable processes that combine AI reasoning, tools, and platform controls.