Quick Start

You’ve got Selu installed and the dashboard is open. Let’s get your first agent up and running. This guide takes about five minutes.

Step 1: Create your account

When you first open the Selu dashboard, you’ll be prompted to create an admin account. Pick a username and password — this is stored locally on your machine, not sent anywhere.

Once logged in, you’ll land on the Chat screen. It’ll look a bit empty right now. That’s about to change.

Step 2: Connect a language model provider

Agents need a language model to think and respond. Selu doesn’t come with one built in — you bring your own.

1. Go to Providers in the sidebar.
2. Click Add Provider.
3. Pick your provider from the list.
4. Follow the setup steps — Selu will guide you through the required fields and run a connection check automatically.

OpenAI

Use models like GPT-4o. You’ll need an API key from platform.openai.com.

Anthropic

Use models like Claude. Grab an API key from console.anthropic.com.

Amazon Bedrock

Use AWS-hosted models. Requires AWS credentials with Bedrock access.

Pico AI Server

Run models on your own hardware with Pico AI Server. No API key needed — supports 300+ models locally.

Tip

If you’re not sure which provider to start with, OpenAI or Anthropic are the easiest to set up. You can always add more providers later or switch to local models via Pico AI Server.

Step 3: Install an agent from the marketplace

Now that Selu has a language model to work with, let’s give it a job.

1. Go to Agents in the sidebar.
2. Switch to the Marketplace tab.
3. Browse the available agents. For your first time, we recommend the General Assistant — it’s a well-rounded conversational agent.
4. Click Install.
5. Selu will configure the agent and assign it a default set of capabilities. You can customize these later.

The agent should appear on the Installed Agents tab within a few seconds.

Step 4: Set up a pipe

Before you can chat, you need at least one pipe — that’s how messages reach your agents. The quickest option is a Web UI pipe, which lets you chat right in the Selu dashboard.

1. Go to Pipes in the sidebar.
2. Click Add Pipe.

3. Select Web UI.
4. Give it a name (for example “General” or “Personal”), pick yourself as the owner, and click Create Web Pipe.

That’s it — your new pipe will show up in the Chat sidebar straight away.

Tip

The Web UI pipe is the fastest way to get started. Later you can add pipes for iMessage, WhatsApp, or Telegram to chat with your agents from your phone.

Step 5: Send your first message

1. Go to Chat in the sidebar and click New conversation.
2. Type a message — something like “Hi! What can you help me with?”
3. Hit send.

Your agent should respond within a few seconds, introducing itself and explaining what it can do. Congratulations — you have a working AI agent running on your own infrastructure.

What just happened?

Here’s what went on behind the scenes when you sent that message:

1. Your message was sent to the Selu server running on your machine.
2. The server passed it to your installed agent, along with its system prompt and available capabilities.
3. The agent forwarded the conversation to your configured language model provider.
4. The model’s response came back through the agent to your chat window.

No third party saw your conversation (beyond the LLM provider, if you’re using a cloud-based model). Everything else stayed on your machine.

Try a few things

Now that you’re set up, here are some things to explore:

Connect a pipe

Go to Pipes in the sidebar and connect Telegram, iMessage, or WhatsApp, so you can chat with your agent from your phone.

Explore capabilities

Check out your agent’s detail page under Agents to see what tools it has access to, and toggle them on or off.

Install more agents

Head back to the Marketplace and try a specialized agent — a writing assistant, research helper, or task planner.

Next steps

Want to understand more about how conversations work and what your agent is actually doing? Continue to Your First Conversation.