Designing for AI Participation
When Sara Du was helping companies build Model Context Protocol servers in 2025, people frequently inquired about running AI agents directly inside Slack. However, moving messages back and forth while providing agents with sufficient context created technical bottlenecks without consuming excessive token budgets.
Investigating this limitation revealed a deeper architectural issue regarding how workplace communication platforms are structured. Legacy apps were built for an era where software played a passive role rather than acting as an active participant in team operations.
The proposed solution eliminates what founders describe as human proxies, where employees are forced to manually relay tasks and findings between AI agents and the rest of the organization.
An optimized system allows agents to take part in shared communication channels where workflow coordination happens naturally. Agents should understand decisions, ask colleagues questions, build upon other agents' work, and escalate to humans when intervention is necessary.
A New Workplace Platform
The startup Ando launched publicly with an application designed specifically for teams consisting of both human employees and autonomous AI workers.
Ando positions itself as a replacement for standard enterprise messaging tools. The platform assigns agents individual identities and inboxes, enabling them to engage in discussions as naturally as human personnel.
The software supports channels, direct messages, group chats, and live calls that can be transcribed for agent review. Agents can independently browse and join channels without explicit tagging, notifying human workers autonomously when important developments arise.
Funding and Market Landscape
To support ongoing development, Ando secured $20 million in pre-seed and seed financing from investors including Accel, Index Ventures, and Emergence.
While major collaboration platforms have introduced agent support—such as Slack's native bot integrations and Microsoft's Copilot within Teams—new entrants aim to challenge incumbents by designing systems from the ground up for agent-native workflows.
Other developer-focused tools, such as Jack Dorsey's Buzz project, are also exploring unified communication environments for humans and AI agents.
Founders believe building a dedicated agent-native platform provides a distinct advantage while legacy software providers attempt to adapt their existing codebases.
Early Adoption and Capabilities
Early demonstrations faced skepticism, with users initially viewing the software as an unconventional messaging interface before grasping the underlying operational differences.
Customer retention improved as organizations utilized the platform for extended periods. Users observed agents performing advanced coordination tasks, such as identifying related discussions in separate channels and spontaneously consolidating teams to propose solutions.
Proponents suggest that agents can manage communication flows efficiently because they process vast volumes of messages in short timeframes.
Ando currently works with small teams across software, real estate, and finance sectors in 15 countries. The newly acquired capital will fund team expansion and computational overhead.
AI agents will enable compact teams to operate at a scale previously requiring hundreds of personnel, handling execution, research, and coordination while humans concentrate on high-level strategy and judgment.
