Introducing Daemora: Your AI That Works While You Sleep

The Problem We Solved

Every AI assistant today has the same fatal flaw: it waits for you.

You type a prompt. It responds. You refine. It responds again. This back-and-forth loop might feel productive, but it's fundamentally broken. You're still the bottleneck. The AI never takes initiative, never follows up, never finishes the job while you're asleep.

We built Daemora to change that.

What Is Daemora?

Daemora is a self-hosted autonomous AI agent that runs on your own machine, connects to your messaging apps (Telegram, Slack, WhatsApp, Discord, Email, Teams, iMessage, Matrix, Twitch, and more - 22 channels total), and executes tasks end-to-end without hand-holding.

Send it a message at 11pm:

"Research the top 5 competitors in the AI agent space, write a comparison doc, and email it to the team by morning."

Go to sleep. Wake up. The report is in everyone's inbox.

Your Data, Your Control

Today, Daemora runs on your hardware - your data stays with you. But we're also building Daemora Cloud for those who want managed compute without the DevOps hassle. Either way, you stay in control.

We built 10 layers of defence-in-depth security because trust has to be earned, not assumed.

What Makes It Different?

60+ built-in tools. File I/O, shell execution, web scraping, browser automation, email, voice calls, document generation, memory, scheduling, and MCP integration.

22 channels. Telegram, WhatsApp, Discord, Slack, Email, Teams, iMessage, Matrix, Mattermost, Twitch, Nostr, LINE, Signal, Google Chat, IRC, Nextcloud, Zalo, Feishu, BlueBubbles, and more. Meet your users wherever they are.

Multi-agent orchestration. Daemora can spawn sub-agents, run them in parallel, and cascade-kill them when done. It plans like a team lead and executes like a full team.

Multi-tenant. Run one instance for your entire team. Each user gets isolated memory, filesystem, API keys, cost limits, and model selection.

Cron scheduling. Set it once, it runs forever. "Every Monday at 9am, pull last week's activity and post a summary to Slack."

Model freedom. Connect to OpenAI, Claude (Anthropic), Google Gemini, or run local models via Ollama. Switch models per user, per task type.

A Weekend Project That Grew Up

Daemora started as a weekend hack - a simple Telegram bot that could run shell commands. Within weeks, it evolved into a production-grade autonomous agent platform with multi-tenant architecture, encrypted vaults, and a full CLI.

It's still a passion project, built by developers for developers. Open source under AGPL-3.0.

Get Started

npm install -g daemora
daemora setup
daemora start

Three commands. Then message your bot. That's it.

The future of AI isn't waiting for prompts. It's agents that work while you live your life.

Welcome to Daemora.