Ando emerges from stealth with $20 million funding to compete with Slack, as AI agents shift from add-on apps to team members

The startup Ando emerged from stealth mode, announcing the raising of $20 million in a seed and pre-seed round led by prominent investors including Axil, Index Ventures and Emergence, to build an enterprise messaging platform designed to bring human employees and AI agents together in a single work environment that aims to compete with platforms such as Slack and Microsoft Teams and replace them entirely.

The idea traces back to Sara Do’s experiment at the institution in 2025, when she was helping companies build servers according to the Model Context Protocol, commonly abbreviated as MCP servers. During that period, companies repeatedly asked how to run AI agents directly from within the Slack platform. However, integration attempts ran into a series of technical obstacles, including difficulty moving messages back and forth, providing agents with appropriate operational context, and avoiding the depletion of costly token budgets.

Current platforms were built for a world that is already outpacing us, where the agent is treated as an installed application rather than a teammate.This conclusion led Sara Do to identify a dilemma she describes as the “human middle-man” phenomenon, where the employee effectively becomes a mere courier delivering the smart agent’s output to other company departments. Do believes the optimal system should allow the agent to participate directly in group conversations where work is coordinated, understand the reasons behind decisions, pose questions to colleagues, build on another agent’s work, and summon the human element only when judgment and evaluation are needed.

The Ando platform gives smart agents their own identities and independent digital mailboxes that let them engage in discussions as smoothly as human colleagues. The platform includes chat channels, one-to-one and group conversations, as well as live voice calls that an agent can transcribe audibly and follow up in text. An agent also has the freedom to browse channels and join them without being mentioned, and to message any employee proactively if it deems it necessary to inform them of a urgent matter, without waiting for prior permissions.

Do notes that competition is not easy in a market where enterprise-communication giants are trying to adapt, as Slack turned its interactive tools into a smart agent, Microsoft deeply integrated its Copilot assistant into Teams and the Office 365 suite, and Jack Dorsey recently launched the Biz platform aimed at merging developers and agents. Although early adopters were cautious, seeing the platform as a less-mature messaging app before perceiving the difference, usage hours rose once models began showing tangible proactive capabilities, such as detecting two separate discussions about the same issue in different channels, prompting the agent on its own to bring the parties together in a unified conversation, explain the context, and propose a decision to address the matter, leveraging its ability to process massive streams of messages in record-breaking times that exceed human capacity.

The platform today serves customers in 15 countries across software, real-estate and finance sectors, and most of its current teams focus on small-scale businesses. The company plans to use the new funding to hire additional talent and cover token consumption, based on a vision that agents will enable extremely small teams to operate at a scale that previously required hundreds of employees, with agents handling execution, research and coordination while humans concentrate on sound judgment, strategy building and guiding next steps.

This shift in communication architecture has direct implications for those leading technical and operational teams in the Gulf, Egypt and the Levant. In regional business environments that are racing today to adopt AI tools and build API-linking servers, developers and product managers spend a long time acting as couriers between models and traditional work channels. Moving to platforms that integrate agents as full members requires rethinking internal coordination skills and authority governance, as the question is no longer limited to the model’s performance but to how ready the organizational structure is to grant agents initiative and address employees directly without descending into uncontrolled message chaos, while strictly balancing privacy requirements and local data-management policies.