One agent trying to serve everyone usually ends up serving no one particularly well.

Your internal HR team starts using the same agent as your customers. Within a week, a customer asks a question and gets a response that references an internal onboarding process. An employee asks about benefits and gets a customer-facing answer about your refund policy. The agent is not broken - it is just trying to serve two completely different audiences with the same knowledge and the same instructions, and it cannot do both well at once.

Separate agents fix this immediately. Each one has its own sources, its own tone, its own rules - and nothing bleeds between them. Outlearn lets you run as many as you need from a single account.

In this article, you'll learn:

• When to use multiple agents
• How to switch between agents
• How to check an agent's live status
• How to create a new agent

When Should You Use Multiple Agents?

A single agent works well for most teams starting out. But as you grow, you may find it useful to split into multiple agents when:

• You have different audiences - for example, a public customer support agent and a private internal agent for your team
• You have different topics that shouldn't mix - for example, a sales agent and a technical support agent
• You want to test a new configuration without affecting your live agent
• You're running agents for different brands or products under the same account

Each agent has its own sources, instructions, and deploy channels - so they stay completely separate.

Switching Between Agents

Your currently active agent is always shown in the top navigation bar. To switch:

1. Click on the agent name in the top navigation bar.
2. A dropdown will appear listing all your agents. The currently active one has a checkmark next to it.
3. Click any agent to switch to it.

Everything in the dashboard - Sources, Actions, Handoffs, Deploy, Analytics, Settings - updates immediately to reflect the agent you've switched to.

Checking an Agent's Live Status

Next to the agent name in the top navigation bar, you'll see small icons showing where that agent is currently deployed and active:

• Globe icon: The agent is live via Chat Widget. Hover over it to see the URL where it's embedded and when it was last active.
• Envelope icon: The agent is live via Email. Hover over it to see the connected email address and when it was last active.

If no icons appear, the agent hasn't been deployed yet or all its channels are turned off.

Creating a New Agent

1. Click on the agent name in the top navigation bar.
2. At the bottom of the dropdown, click Create new agent.
3. Fill in the following:
   • Agent Name (required)
   • Description (optional) - A short note about what this agent is for
   • Agent Icon (optional) - PNG, JPG, or SVG up to 5MB. Shown in the agent switcher.
4. Click Create Agent.

Your new agent opens immediately with an empty dashboard. From here, set up its Sources, configure its Settings, and deploy it when it's ready.

A new agent starts completely blank - no sources, no actions, no handoffs. You'll need to configure it from scratch. For a step-by-step guide, see Onboarding Walkthrough: Setting Up Your First Agent.

Best Practices

• Give every agent a clear, descriptive name.
• Use the Description field to note the agent's purpose and audience. It takes 10 seconds and saves confusion later.
• Don't create a new agent just to test a change - use the Preview panel to test changes on your existing agent without affecting live users.
• Keep your agent list tidy. Unused or test agents can be deleted from Settings → Danger Zone when you no longer need them.