Artificial intelligence has rapidly entered the workplace, but most companies still use it in isolation. Employees prompt chatbots individually, receive answers privately, and move on, leaving valuable context and knowledge trapped inside disconnected conversations. While productivity may improve for individuals, organisations often fail to capture collective learning or long term operational intelligence. Agentic AI company Dust believes the next phase of AI adoption will move beyond isolated assistants toward collaborative AI systems shared across entire organisations, and investors are backing that vision with fresh capital.
Dust has raised $40 million in a Series B funding round led by Abstract and Sequoia, with participation from Snowflake and Datadog. With the latest financing, the company has now raised more than $60 million in total funding.
Building an Operating System for AI Agents
Dust describes itself as an operating system for AI agents designed to help organisations deploy, orchestrate, and govern fleets of specialised AI systems working alongside employees.
Rather than functioning as isolated chat assistants, Dust’s platform allows businesses to create AI agents connected directly to company knowledge, tools, workflows, and operational context. The platform is built around a shared collaboration layer where humans and AI agents operate together inside the same workspace using shared projects, conversations, files, notifications, tasks, and workflows.
The company refers to this model as “multiplayer AI,” where humans and AI systems collaborate continuously rather than interacting through isolated one to one prompts.
According to Gabriel Hubert, the real transformation in AI will not come solely from increasingly powerful models but from systems that allow humans and agents to share context, goals, and operational capabilities in real time across entire organisations.
Connecting AI to Enterprise Knowledge
Dust’s platform includes an intelligence layer capable of integrating with more than 100 enterprise data sources and workplace tools. This allows AI agents to access organisational context while interacting with the software systems employees already use daily.
The company says its infrastructure enables agents to not only retrieve information but also take operational actions within enterprise environments. Built in memory systems and reinforcement loops further allow agents to improve over time by learning organisational preferences and recommending workflow optimisations.
Enterprise governance features are also central to the platform. Dust includes permission controls, audit trails, usage monitoring, cost tracking, and analytics tools designed for large scale enterprise deployments. The platform is SOC 2 Type II certified, GDPR compliant, supports both European and US data residency requirements, and contractually guarantees that customer data is not used to train AI models.
Rapid Enterprise Adoption
Dust says the platform is now used by more than 3,000 organisations globally, including many large enterprise customers. More than 300,000 AI agents have already been deployed across the platform, while the company reports weekly active usage rates exceeding 70 percent among customers and zero customer churn during 2025.
Investors say these adoption metrics demonstrate that enterprise AI is beginning to move beyond experimental pilots toward becoming operational infrastructure inside companies.
Konstantine Buhler said most enterprise AI systems today remain “single player” experiences where individual users interact with isolated tools. According to him, Dust is building a collaborative AI environment where humans and agents work together across the organisation while sharing operational context and institutional knowledge.
Experienced Founders With AI and Product Backgrounds
Dust was founded by Gabriel Hubert and Stanislas Polu, who previously co founded data analytics company TOTEMS before its acquisition by Stripe in 2014.
Following the acquisition, both founders spent several years at Stripe before Polu later joined OpenAI as a research engineer working on AI reasoning projects alongside researchers including Ilya Sutskever. Hubert later became Chief Product Officer at Alan.
The company says the new funding will support three major priorities: building AI agents capable of continuously learning and improving automatically, developing collaboration systems where humans and AI act as equal contributors, and expanding enterprise infrastructure for governance and orchestration at scale.
