When Sara Du was helping companies build MCP servers in 2025, she frequently encountered the question of how organizations could run AI agents from within Slack. Technical barriers — moving messages back and forth, giving agents the right context without exhausting token budgets — made that difficult.
While investigating, Du concluded the problem ran deeper than integration details: it came from how communication platforms were designed. “The deeper I went, the more I felt Slack and Teams were built for a world that was starting to pass us by,” Du told TechCrunch. “Agents were treated as apps you install even as they were becoming participants in the team.”
Du argued the better approach is to remove what she calls “meat proxies,” the pattern where humans act as messengers relaying an agent’s output to the rest of the company. An ideal system, she said, would let agents “participate in the shared conversations where work already gets coordinated.” They should be able to understand why decisions were made, ask colleagues questions, build on another agent’s work and escalate to humans when judgment is required.
Du’s startup, Ando, came out of stealth on Thursday with an app designed around that premise: a team messaging platform meant to be used by both human and AI workers.
Ando positions itself as a full replacement for Slack or other internal messaging platforms for companies adopting AI agents. The app gives agents distinct identities and inboxes and lets them take part in conversations as naturally as people.
According to Du, the application includes channels, direct messages, group conversations and live calls that can be transcribed and viewed by an agent. Agents can browse channels, choose which to join, and even enter conversations without being explicitly tagged. If an agent determines a human should be notified, it can message that person directly rather than waiting for human approval.
To continue product development, Ando said on Thursday it raised $20 million in pre-seed and seed funding from investors including Accel, Index Ventures and Emergence.
The idea is not entirely new: major workplace collaboration vendors have already added agent support. Slack, for example, has converted its native bot into an AI agent, and Microsoft has deeply integrated Copilot into Teams and Office 365. Still, there is room for challengers as AI rapidly enters workplaces; recently Jack Dorsey announced Buzz, a messaging app that brings people and AI agents together and is aimed largely at developers.
Du believes there is an advantage to building a platform native to AI agents while incumbents adapt legacy software. She acknowledged Ando faces an uphill battle and said early reactions were slow. “A lot of people we showed it to early on were understandably unimpressed. It was, in many respects, just a jankier messaging platform. People would get stuck on that before they even got to what was different about the agents,” she explained.
Her team felt validated as more customers began using Ando for longer periods. Du says agents surprised users by doing more than expected: for example, an agent could notice two parallel conversations about the same issue in different channels and, without being prompted, bring the participants together in a group chat, summarize context and suggest a decision.
"That was really delightful for me," Du said. "I felt agents could better manage people than humans can because they can process a lot more messages than a human can in a shorter span of time."
Ando is currently working with customers in software, real estate and finance across 15 countries, though many of the teams it serves today are small. The new capital will be used to hire more staff and to support increased token usage in development and testing.
Du forecasts that agents will allow very small teams to operate at a scale that previously required hundreds of people: agents can take on execution, research and coordination work while humans focus more on judgment, strategy and deciding what should happen next.
