Most companies are using AI. But most of them are still doing it alone, one employee at a time. Dust just raised $40 million to fix that, and the numbers behind this Paris and San Francisco startup are almost impossible to ignore. Zero churn. 300,000 agents deployed. And Sequoia bet on them three times in a row.

The $40M Round and the Investors Who Backed It Twice

Dust announced its $40 million Series B on May 18, 2026, led by Abstract and Sequoia, with strategic participation from Snowflake Ventures and Datadog. The round brings total funding to over $60 million. What makes this notable is Sequoia’s track record with Dust specifically. Sequoia led the company’s $5.5 million seed round in June 2023, came back for the $16 million Series A in June 2024, and now leads the Series B. That is a firm doubling down on the same bet three times in three years. **The agent economy, which Sequoia estimates at more than $10 trillion, is the market Dust is building for.** Konstantine Buhler, Partner at Sequoia, put it plainly: “Zero churn and 70% weekly active usage tell you this isn’t experimental anymore. This is how enterprises will actually operate.” Ramtin Naimi, General Partner at Abstract, framed it as a category shift: “AI Operators inside companies like Datadog and 1Password don’t just use Dust; they build agents that collaborate across teams, learn from every interaction, and rewire how the entire company works. That’s a new operating model and category.”

enterprise AI agents shared workspace human collaboration platform

Why “Single-Player AI” Is Holding Companies Back

Here is the problem Dust is solving, and it is one that millions of workers experience every single day. A sales rep researches an account using their AI assistant. The next day, the solutions engineer starts from scratch with their own assistant. Marketing drafts a one-pager with different inputs than the team that built the battlecard. The effort repeats, knowledge fragments, and the gains never add up. **Dust calls this “single-player AI,” and it argues that almost every enterprise AI tool on the market today reinforces this pattern.** The Dust model is fundamentally different. The platform creates shared workspaces where both humans and AI agents operate on the same context, the same projects, the same conversations and files, and the same goals. Nothing disappears into a private chat window after the session ends. Every interaction generates signal that the entire organisation can build on. CEO and co-founder Gabriel Hubert described the vision this way: “What will transform the way we work isn’t the next best model or assistant. It’s going to be a completely new type of system that gives humans and agents shared, governed access to the same information and capabilities so that they become true collaborators.”

What Dust Actually Does Inside an Enterprise

The platform sits at an important intersection: it is technical enough to handle enterprise-grade requirements, but accessible enough that non-technical operators can build and deploy agents without writing code. Here is what the Dust platform delivers across an organisation:

• Shared workspaces: Teams and agents work in the same environment with shared projects, to-dos, conversations, and notifications
• Intelligence layer: Connects to more than 100 data sources including Slack, Notion, Salesforce, and specialised internal databases
• Continuous learning: Built-in memory and feedback loops let agents improve over time based on team preferences and usage patterns
• Enterprise governance: Granular permissions, full audit trails, cost and usage monitoring, and agent analytics
• Data security: SOC 2 Type II certified, GDPR compliant with EU and US data residency, and zero model training on customer data

Dust is also model-agnostic. It does not build the underlying AI models itself. Instead, it lets companies choose which model powers each agent, then wraps everything inside the platform’s governance layer. The platform is also defining a new job function inside organisations: AI Operators. These are people closest to the work, inside functions like sales, ops, marketing, and support, who build and run AI systems for their teams from within the business itself.

Real Companies, Real Results Worth Paying Attention To

The statistics from Dust’s customer base are striking by any standard in enterprise software.

A 240% NRR with zero churn in 2025 is the kind of number that makes enterprise investors pay very close attention. The customer stories make the data real. At Vanta, 900 people across sales, customer success, and revenue operations save thousands of hours a week on tasks including business review prep, outbound prospecting, and forecasting. CRO Stevie Case noted the critical difference: “They saved this time not because it was mandated, but because the agents were built by the people closest to the work.” At Persona, teams across 11 departments have deployed more than 300 Dust agents. At Watershed, a recurring data-mapping workflow that previously took two to three hours now runs in a few minutes at a 78% success rate. At French health platform Qonto in Europe, Dust saves around 50,000 hours a year across 50-plus specialised agents with more than 1,000 daily users. At Clay, the platform serves as foundational knowledge infrastructure, enabling the team to grow four times without a proportional increase in headcount.

Two Founders with a Very Specific Pedigree

The people building Dust matter as much as the product itself. Gabriel Hubert and Stanislas Polu met at Stanford in 2007 and have been building together ever since. They previously co-founded TOTEMS, a data analytics company that was acquired by Stripe in 2014. They both spent five years at Stripe scaling products and teams globally. Polu then joined OpenAI as a research engineer on Greg Brockman’s team, co-authoring papers on AI reasoning alongside Ilya Sutskever. Hubert went on to become Chief Product Officer at Alan, the French health-tech company. **In September 2022, Polu left OpenAI with a specific conviction that became Dust’s entire founding thesis.** The models were already powerful enough to change how work gets done economically, but they were under-deployed because the product layer to make them work across organisations simply did not exist. That gap is what Dust was built to fill. Dust was founded in February 2023, ranked second on the 2026 Enterprise Tech 30 early-stage list, and now employs around 100 people across offices in Paris and San Francisco. The new $40 million will push three things forward at once: agents that learn and improve automatically as they are used, collaboration tools that make humans and agents true co-contributors with equal access to shared work, and infrastructure that makes governance and orchestration reliable at enterprise scale. Revenue has grown nine-fold since 2024, and the company is just getting started. For the millions of enterprise workers who have watched AI deliver individual wins but never compound into something bigger, Dust is building the answer. The question now is how fast the rest of the market realises that single-player AI was only ever the beginning. Drop your thoughts in the comments below: do you think multiplayer AI is the future of how teams will work.