From Single-Player Chatbots to Enterprise AI Agents at Scale
Enterprise AI has been dominated by “single-player” experiences: one person, one chatbot, one prompt, and very little shared learning across teams. Dust’s trajectory shows that phase is ending. The company positions itself as an operating system for enterprise AI agents, enabling organisations to deploy, orchestrate, and govern fleets of specialised agents that work alongside employees. More than 300,000 enterprise AI agents have been deployed on Dust’s platform across over 3,000 organisations, a scale that is difficult to reconcile with mere experimentation or isolated pilots. These agents are not generic assistants; they are embedded into company tools and knowledge, designed to execute workflows, collaborate with colleagues, and act on up-to-date context. This shift from occasional chatbot usage to systematic AI agent deployment is a strong signal that enterprises are moving toward AI as part of their operational fabric rather than a side experiment.
Usage Metrics That Look More Like Core Software Than Experiments
Adoption volume alone can be misleading, but Dust’s engagement and retention metrics indicate deep embedment into daily work. The platform reports 70% weekly active usage across its customer base and zero churn in 2025. These are not the numbers of a novelty tool; they resemble a core business system whose removal would break workflows. In a crowded field of AI platforms, such metrics are strong evidence of product–market fit around enterprise AI agents. Customers are not only deploying agents; they keep coming back to use them each week, suggesting that agents are handling recurring processes rather than one-off trials. For decision-makers, this level of stickiness implies that AI agents are starting to own durable parts of the value chain—support operations, knowledge workflows, internal tooling—marking a shift from proof-of-concept thinking to long-term dependency on AI-powered systems.
Why Investors Are Backing Multiplayer AI Platforms as Infrastructure
Dust’s USD 40 million (approx. RM184 million) Series B, led by Abstract and Sequoia with participation from Snowflake and Datadog, underscores a growing belief that multiplayer AI platforms will form a key layer of enterprise infrastructure. Investors are not just betting on a better chatbot; they are backing an operating system where humans and agents share context, projects, notifications, and a common compute environment. This multiplayer AI platform model connects to over 100 data sources, integrates into existing tools, and wraps everything in enterprise governance—granular permissions, audit trails, and usage analytics. That combination looks less like an app and more like a foundational AI layer. The funding validates the view that the missing piece in enterprise AI is not more powerful models, but infrastructure that makes it safe and repeatable to deploy and coordinate agents across an entire organisation.
Operationalising AI: From Experiments to a New Operating Model
The velocity and structure of Dust’s deployments point to a broader operational shift. Over 300,000 agents running across thousands of organisations implies that teams are standardising on AI agents for repeatable workflows, not just ad hoc experiments. Dust’s shared workspace lets humans and agents collaborate on the same projects and conversations, while an intelligence layer and built-in memory allow agents to learn preferences and improve through reinforcement loops. Enterprise governance and compliance features, including SOC 2 Type II certification and contractual guarantees that customer data is not used for model training, further indicate readiness for mission-critical use. Together, these elements define an emerging operating model: AI agents as persistent collaborators embedded in everyday work. For enterprises, the question is rapidly changing from “Should we experiment with AI?” to “How do we architect, govern, and scale AI agents as a core part of how we operate?”