Ando Aims to Challenge Slack With Messaging App Built for Human-Agent Collaboration

Ando Wants to Be Slack for Teams Where AI Agents Are Actual Coworkers

A startup called Ando is building a team messaging platform that treats AI agents as full participants rather than tools. The company emerged from stealth this week with $20 million in funding from Accel, Index Ventures, and Emergence Capital .

The problem it's trying to solve

Ando founder Sara Du previously built enterprise integration tools and kept running into the same issue. Companies wanted AI agents to work directly inside Slack, but the experience was clunky. Messages had to be relayed back and forth, context got lost, and humans ended up acting as what Du calls "meat proxies" — messengers between the agent and the rest of the team .

Her argument is that Slack and Teams were designed for a world where every participant is human. Agents get treated like apps you install, not coworkers you talk to. Ando is built from scratch around a different assumption: agents deserve their own identities, inboxes, and permissions .

How it actually works

The platform looks familiar at first. There are channels, DMs, group chats, and live calls. The difference is that agents can browse channels, join conversations without being tagged, and proactively message people when they think something needs attention .

Agents are also agent-agnostic. Teams can bring their own tools like Codex, Claude, or Grokbot, or use Ando's hosted version. The platform keeps persistent context from conversations, so an agent understands why a decision was made, not just the latest message .

Du describes a scenario where an agent notices two different channels discussing the same problem, pulls everyone into a group chat with the relevant context, and even suggests a decision. That kind of coordination, she says, is something agents can do better than humans because they can process far more messages in less time .

The uphill battle

Ando is entering a crowded space. Slack has already turned its native bot into an AI agent, and Microsoft has deeply integrated Copilot into Teams. Jack Dorsey recently announced Buzz, a messaging app for humans and AI agents aimed more at developers .

Du acknowledges that early demos were underwhelming. People saw "just a jankier messaging platform" before they understood what made the agents different. But she says teams started sticking around once they experienced agents proactively connecting dots .

The company is currently working with teams in software, real estate, and finance across 15 countries, though most are small. Ando itself has completely stopped using Slack .

What I think

The "new Slack" pitch is bold, but the underlying observation is sharper than the marketing. Every major collaboration platform is bolting AI agents onto software that was never designed for them. Ando's bet is that starting fresh, with agents as first-class members, creates something fundamentally different.

The most interesting detail is the pricing model. Ando charges by human seats, not by agent activity. That removes the incentive to ration agent participation, which is exactly the kind of friction that makes current tools frustrating .

The real question is whether "agent-native" matters enough to overcome Slack's network effects. Switching costs are brutal. Ando is targeting small teams first, which makes sense — they have less to lose and fewer entrenched workflows. Whether that traction scales to larger companies is the part that remains unproven.

If agents really do become coworkers rather than tools, someone will build the communication layer for that world. Ando is making an early, focused bet that it can be that someone.