Lilo brings an open source AI assistant to Telegram

Lilo is an open source personal AI assistant that runs in Telegram and other channels, including email, WhatsApp, web, and mobile apps. Its creator says it is model-agnostic, keeps memory in a persistent workspace, and can use tools such as web search and browser control.

Lilo brings an open source AI assistant to Telegram

An open source personal AI assistant called Lilo now lives in Telegram, with support for email, WhatsApp, a website, and a mobile app as well. Its creator introduced the project in a Show HN post and said Lilo is built as a model-agnostic agent with memory, tools, scheduled tasks, and a persistent workspace.

According to the post, Lilo can handle a range of personal tasks through chat and voice. The creator said users can send photos of food for calorie tracking, leave a voice note to create a todo item, ask for reminders around sports games, get article summaries read aloud, and forward receipts or invitations for later handling.

⚡ New to this?

This matters because it shows how personal AI assistants are moving from simple chatbots to tools that can keep memory, run tasks on a schedule, and work across several apps. Telegram is a messaging app, and a persistent workspace means the assistant can remember context and keep useful files or mini-apps over time. For non-experts, the big change is that the assistant is not just answering questions - it is being used to manage real daily chores.

🦞 OpenClaw angle

If you build self-hosted AI agents, treat Lilo as a reminder to separate the chat interface from the agent backend. A remote, always-on workspace makes it easier to run scheduled jobs, keep state, and accept multiple input channels without tying everything to one local machine. If you are designing your own agent, add support for email forwarding and voice notes early, since those are practical high-volume inputs for real users.

The system also supports more structured planning. In the examples shared by the creator, Lilo can help schedule a meeting, suggest locations, and then send a reminder later so the user leaves on time.

The assistant is not limited to text responses. The creator said Lilo can use tools such as web search and a browser user, and it can build HTML-based apps that live inside its workspace.

That workspace is persistent, meaning the assistant can keep state over time rather than treating each interaction as a one-off request. The creator also said the setup includes scheduled tasks, which lets Lilo act on reminders or recurring events without the user needing to prompt it each time.

The project differs from the creator’s other assistants, OpenClaw and Hermes Agent, in a few ways. According to the post, Lilo runs on a remote machine or in the cloud instead of on the user’s local machine, which the creator says keeps local data safe and makes the assistant available 24/7.

The creator also described Lilo as more visual than text-only assistants. It ships with apps such as a todo list that users can interact with through a GUI on mobile and web, not just through chat messages.

Telegram appears to be the main interface for the assistant, and the creator said that integration is broad enough to handle replies, voice notes, reactions, and other common message types. The post says email is also useful because users can forward receipts and invites for Lilo to process later.

The project is open source and the creator invited feedback from the HN community. The GitHub repository for Lilo was shared in the post.

Source: HN Show HN ↗

More from OpenClaw News