Dust: Raises $40 Million to Build the “Multiplayer” Operating System of the Enterprise AI Era
Dust is developing what it describes as the "Operating System for AI Agents," a platform that allows enterprises to create, manage, and orchestrate fleets of specialized AI agents that collaborate with human teams.
Dust announced new funding of $40 million in a Series B round, led by Abstract and Sequoia Capital, with participation from Snowflake and Datadog. With this new investment, the company has raised over $60 million in total.
Dust is building a platform that allows businesses to create, manage, and orchestrate fleets of specialized AI agents working alongside human teams, leveraging corporate data and the tools organizations already use.
According to the company, the core problem with today’s AI assistants is that they operate in isolation: an employee interacts with a chatbot, receives an answer, and all the context is lost within a private chat window. This creates individual productivity, but fails to build cumulative knowledge at the organizational level.
Dust aims to change this model by introducing the concept of “multiplayer AI.” In this setup, humans and AI agents collaborate within the same workspace, sharing projects, notifications, files, workflows, and common context.
The platform connects to more than 100 data sources and integrates with existing business tools, allowing agents to leverage corporate knowledge and take action. At the same time, it features memory mechanisms and reinforcement loops so that AI agents gradually improve through use.
A crucial part of Dust’s proposition is its enterprise governance layer, which includes:
- Granular permissions
- Cost and usage monitoring
- Audit trails and analytics for AI agents
The company states it is SOC 2 Type II certified, GDPR compliant, and supports data residency in both Europe and the US. Importantly, they emphasize that the AI models are not trained on customer data.
The platform’s growth appears to be exceptionally rapid. Dust claims it is being used by more than 3,000 organizations, with over 300,000 AI agents already deployed through the platform. Furthermore, the company reports a 70% weekly active usage rate and zero churn throughout 2025.
The company was founded by Gabriel Hubert and Stanislas Polu, who met at Stanford in 2007. The two founders previously co-founded TOTEMS, a data analytics company acquired by Stripe in 2014, and subsequently worked at Stripe for five years.
Later, Polu joined OpenAI as a research engineer, working on Greg Brockman’s team and co-authoring papers on AI reasoning alongside Ilya Sutskever. He left OpenAI in 2022, convinced that while AI models had reached a level capable of economically transforming businesses, the appropriate product layer for mass adoption was still missing.
Dust plans to use the new funding to develop:
- AI agents that automatically improve through ongoing use.
- New collaboration tools designed for tighter integration between humans and agents.
- Enterprise infrastructure that enables predictable governance and orchestration at scale.
