Dust is launching a Series B of $40 million to expand its multiplayer, enterprise-grade AI platform that enables collaborative agents shared across an organization, addressing the limits of single-player AI.

The platform connects to more than 100 data sources and integrates with tools like Slack, Salesforce, and Google Drive, allowing agents to access up-to-date company information and learn from real team workflows through memory and feedback loops.

The funding is led by Sequoia and Abstract, with participation from Datadog and Snowflake.

Founder Hubert says profitability is not the near-term goal; the raise is aimed at accelerating R&D and growth, particularly in the US.

Notable customer outcomes include 4x GTM team scaling at Clay, shorter onboarding for Doctolib, and faster cross-functional workflows like RFPs at Persona.

Dust surpassed $20 million in annual recurring revenue earlier this year and is not yet profitable.

Dust emphasizes enterprise-grade governance with granular permissions, cost tracking, audit trails, SOC 2 Type II, GDPR compliance, data residency in EU and US, and a contractual guarantee that customer data is not used to train models.

The Series B will advance three priorities: agents that improve with use, bidirectional human-agent collaboration primitives, and enterprise-scale orchestration and governance; run-rate revenue was not disclosed.

Dust reports traction of serving 3,000+ organizations, with 70% weekly active usage and zero churn in 2025, plus deployment of over 300,000 agents.

Dust’s net revenue retention reached 240% in 2025, reflecting growing multi-department adoption across organizations.

Notable customers include Vanta, Clay, Profound, Persona, and Doctolib, with Doctolib using Dust to centralize internal AI across thousands of employees.

The Series B will fund further US expansion and R&D, highlighting that US deployments are already at scale and will be decisive for enterprise AI adoption.