Agentic AI
Ando Launches Slack Alternative Built for Teams Where Humans and AI Agents Collaborate
A new startup is challenging workplace communication incumbents with a messaging platform designed from the ground up to let AI agents participate as equal team members rather than external tools.
TechCrunch — AI
Sara Du encountered a recurring problem while helping organizations implement AI agent infrastructure in 2025: workers wanted to deploy these agents directly within Slack, but the technical barriers were substantial. Integrating AI systems required complicated workarounds to handle message routing, context provision, and token management—each solution felt like a patch rather than a true integration.
This observation led Du to a larger insight about the fundamental architecture of existing workplace platforms. "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. Current systems, she realized, treat AI agents as applications rather than as active contributors to team workflows.
A New Model for AI Integration#
Du's critique centers on what she calls "meat proxies"—the burden on human workers to interpret and relay agent outputs back to the team. "The human effectively becomes the messenger between the agent and the rest of the company," she explained.
Instead, her vision reimagines how agents participate in team collaboration. Rather than operating as isolated tools, agents should inhabit shared conversation spaces where work unfolds. This would enable them to understand decision context, pose questions to colleagues, build upon other agents' contributions, and escalate issues requiring human judgment.
Ando's Architecture#
Ando, which exited stealth on Thursday, translates this concept into a full messaging platform designed to replace Slack for AI-forward organizations. The system grants agents their own identities, inboxes, and conversation capabilities that mirror human participation.
The platform supports:
- Channels and direct messaging for structured and one-on-one communication
- Group conversations that pool multiple participants
- Live calls with transcription and agent-accessible recording
- Autonomous agent behavior—joining channels and conversations without explicit invitation, or independently messaging humans when warranted
Agents can independently browse channels, select which ones to monitor, and even initiate multi-party discussions when they identify overlapping problems across different threads.
