User Guide Overview

Selu is a personal agent platform that lets you run AI agents on your own infrastructure — no coding required. You stay in control of your data, your models, and how your agents behave.

How it works

At its core, Selu connects channels (the places you chat) to agents (the AI that responds). Messages flow in from a channel, get routed to the right agent, and the agent replies using the language model you’ve configured.

Everything runs inside Docker containers on your own hardware or cloud server. Agents communicate with Selu through gRPC, which keeps things fast and reliable behind the scenes.

Key concepts

Channels

The interfaces where you talk to your agents — web chat, iMessage, or Telegram. Connect one or all of them.

Agents

AI-powered helpers you install from the marketplace. Each agent runs in its own container and can have unique skills.

LLM Providers

Choose the language model behind your agents — Anthropic Claude, OpenAI GPT, Amazon Bedrock, or a local model via Ollama.

Security

Encrypted credentials, tool approval policies, and full control over what your agents can and cannot do.

What you’ll find in this guide

This user guide walks you through everything you need to get the most out of Selu:

• Channels — Set up web chat, iMessage, and Telegram so you can talk to your agents wherever you are.
• Agents — Install agents from the marketplace, configure sessions, and set up event-driven notifications.
• Personality & Memory — Customize how your agent sounds and what it remembers about you.
• LLM Providers — Connect to your preferred AI model provider.
• Security — Manage credentials safely and control which tools agents are allowed to use.
• Self-Hosting — Run Selu with Docker, configure environment variables, and keep everything up to date.

Next steps

If you haven’t installed Selu yet, head over to the Getting Started guide first. Otherwise, pick any section above and dive in.