Flow Engineering Raises $50M for Hardware AI

Flow Engineering has raised a $50M Series B at a $750M valuation to build AI agents for hardware development. The September 30, 2026 round was co-led by Antonio Gracias of Valor Equity Partners and Gavin Baker of Atreides Management, with Sequoia Capital, Human Capital, Evantic, SV Angel, Odyssey, EQT and several individual investors participating.

The money is important because hardware teams cannot borrow software's speed by simply adding a coding assistant. A change to a physical product can travel through requirements, CAD, simulation, embedded code, tests and regulatory constraints before anybody knows whether the whole system still works. Flow Engineering is building a connected system of record where agents can follow those changes, identify downstream effects and check requirements and test coverage.

What Flow Engineering Raised

Flow's official announcement says the $50M Series B values the San Francisco company at $750M. Valor Equity Partners founder Antonio Gracias and Atreides Management managing partner Gavin Baker co-led the financing. Existing investor Sequoia Capital participated after leading Flow's $23M Series A, alongside Human Capital, Evantic, SV Angel, Odyssey and EQT.

Hugging Face co-founder Thomas Wolf, Mercedes-Benz CIO Jonas von Malottki and Formula 1 world champion Nico Rosberg also invested. Roelof Botha joined Flow's board as an independent director and made a personal investment. Flow previously announced an $8.5M Seed in 2022 and a $23M Series A in 2025, bringing disclosed funding across those rounds to $81.5M.

Why Hardware Needs a Different AI Stack

Software teams can change code, run tests and roll back a release. Hardware programs carry a different consequence structure. Mechanical, electrical and software decisions share interfaces, safety constraints and verification obligations. A missed dependency may not appear until a prototype fails, a supplier builds the wrong part or a regulated program cannot prove that a requirement was satisfied.

Flow connects requirements, CAD, simulation, code and test data. Its systems-engineering agents monitor changes, perform impact analysis, flag conflicts and identify gaps in requirement or test coverage. The company describes the product as an AI harness for hardware engineering, with controls designed to let frontier models work against sensitive engineering data inside live programs.

That positioning separates Flow from a generic engineering copilot. The commercial product is not merely generated text or code. It is a traceable operating layer for deciding whether a complicated physical system still agrees with itself after engineers change it.

Customer Adoption Is the Strongest Evidence

Flow says thousands of engineers at Rivian, Anduril, General Motors, Joby Aviation, Astranis and Radiant Industries use its platform. Since the Series A, the company reports adding General Motors PPU, Rivian and Volkswagen's joint venture RV Tech, Anduril, Stoke Space, Intuitive Machines and Pacific Fusion.

The most specific adoption figure comes from Rivian. Flow says usage expanded from 40 to 1,500 users in seven months and now generates millions of API calls each week. Those are company-reported metrics, not independently audited results, but they explain why investors are treating the product as infrastructure rather than another experiment around AI-assisted design. TechCrunch also reported that Flow is used across automotive, aerospace and defense programs.

What the Series B Changes

Flow plans to expand review, branching and evaluation capabilities, hire across AI and systems engineering, grow its sales team and pursue FedRAMP authorization and other certifications for regulated customers. Each priority points at the same market requirement: autonomous engineering work becomes valuable only when customers can review it, control it and demonstrate why the result should be trusted.

Founder and CEO Pari Singh started Flow after working as a mechanical engineer and building tools for rocket-engine development. The company moved its headquarters from London to San Francisco as it pursued the American aerospace, autonomy, defense and energy market. There is no separately verified current CTO in the reviewed primary and reputable secondary sources, so this article does not invent one.

The Market Consequence

Flow is competing for the layer where hardware organizations coordinate complex development, a category that includes requirements-management, product-lifecycle and engineering-data systems. Its argument is that those systems become more useful when AI can reason across the relationships among requirements, models, tests and design changes instead of waiting for people to reconcile them manually.

The $750M valuation prices in a large version of that argument. Flow still has to turn company-reported adoption into durable enterprise expansion, complete the security and compliance work demanded by regulated programs and prove that faster iteration does not trade away engineering rigor. The Series B gives the company more room to build those controls while its customers are already using the product inside vehicles, spacecraft and other physical systems where the cost of a missed dependency does not stay inside a demo.