We started NatorOS because the gap between what AI agents could plausibly do and what they actually did inside a company was enormous, and the gap was not a model problem. It was an operating-system problem.
Today we are announcing $80M in Series B funding to keep building that operating system. The round was led by Founders Fund and includes participation from our existing investors at Sequoia and Index, as well as a group of operators we have learned from for two years.
Two things became clear in our first year of customers:
NatorOS sits between those two needs. The operator authors a workflow in plain language. The engineer refines the boundaries and approvers. The runtime owns memory, permissions, evals, audit, and replay. Together it produces an AI worker that a CIO can sign off on.
Three places. First, more forward-deployed engineers, our most senior hires, who pair with each customer through the first two weeks. The bottleneck on our growth right now is them. Second, the runtime layer: longer-running workflows, better failure modes, deeper observability. Third, opening offices in London and Singapore so we can serve customers in Europe and APAC with the same hands-on care we give North American teams.
If you are running an operations team and the idea of an AI worker that you, specifically, would trust sounds interesting, book a demo. We will tell you whether your first workflow is ready in fifteen minutes.