A New Chat App Where AI Joins In Like a Coworker
Ando came out of stealth on September 24, 2026 with a workplace chat app where people and AI helpers work together.
The company was founded by Sara Du, according to TechCrunch. It disclosed $20 million in early pre-seed and seed funding from Accel, Index Ventures and Emergence.
Ando says it can fully replace Slack or other work chat tools. It is not an add-on bot or side helper. It is meant to be the main home for work conversation.
The key idea is that each AI helper gets its own name and inbox, like a coworker. You can message it directly. You do not need special shortcut commands or automatic app alerts to call it.
Ando has normal chat tools. It has channels, direct messages, group chats and live calls. Calls can be written out as text so an AI can read them later. Spoken meetings then work the same way as typed chats.
The biggest change is independence. AI helpers can look through channels, pick which ones to join, and join chats without being tagged. Presence is opt-in by the agent. Discovery is agent-driven.
The broader context here is a change in how chat treats AI. Older bots were tools kept in a drawer until called, with narrow jobs. Ando treats agents as members with accounts who can move around on their own. For companies with big Slack or Teams setups, that matters. A called tool stays contained. A member needs account management.
For daily use, the inbox helps. Task lists, mentions, unread marks and threads give managers a normal way to assign work to AI and check what it saw. It also raises normal system questions. How message order is kept when people and AI post together. How rules work when an AI joins a channel that later gets private history. How storage, legal search and access removal apply to an inbox owned by software.
In my view, the bet is clear. If AI is going to do steady work, it cannot wait only in ticket lines or side windows. It needs to be where information first appears, including casual talk that never becomes a formal task. Letting it listen, join and remember captures that value. The tradeoff is control. Joining without a tag shifts the default from direct invite to background listening, and buyers will want clear answers on visibility, consent and containment before they allow it.
Worth flagging for technical buyers is the upkeep this adds. Each AI needs account setup and removal, safe keys, speed limits and logs tied to its name. Each inbox needs sorting and overload control. Self-joining needs rules for who can block an AI, who can see what it read, and how to rebuild its memory after a problem. None of that is exotic. It is basic account care and monitoring for a new kind of user. Teams that already manage automatic accounts for software tests, service accounts and data jobs will know the pattern.
The optimistic case is plain. One shared chat place for people and AI could mean fewer handoffs and less lost context. My kids treat group chat as where plans get made, not email. Workplaces learned that years ago. If AI can live there with a clear name and reviewable history, teamwork gets easier without asking people to talk somewhere else.
