Foxglove Raises $40 Million Series B to Expand Data Platform for Physical AI and Robotics

Foxglove, a San Francisco–based startup developing a data and observability platform for the emerging field of Physical AI, has raised $40 million in Series B funding, marking a major milestone in the company’s growth. The round brings Foxglove’s total funding to more than $58 million since its founding in 2021 and reflects rising investor interest in infrastructure that supports robotics and autonomous systems operating in real-world environments.

The Series B round was led by Bessemer Venture Partners, with participation from existing investors Eclipse Ventures and Amplify Partners. The continued backing from prior investors alongside a new lead underscores confidence in Foxglove’s vision to become a foundational data platform for Physical AI development.

Foxglove builds tools that help robotics and autonomous system developers collect, manage, visualize and analyze massive volumes of multimodal data generated by machines operating in the physical world. These systems produce complex streams of information, including sensor data, video, audio, 3D perception and time-series signals, which must be interpreted accurately to ensure reliable performance. Foxglove’s platform enables teams to debug behavior, validate models and iterate more quickly across development, testing and deployment.

The newly raised capital will be used to expand Foxglove’s platform across the full data lifecycle, including recording, storage, search, analysis and visualization. The company plans to invest heavily in scaling support for petabyte-scale datasets, improving usability for large engineering teams and broadening the platform’s applicability across a wider range of robotics use cases. Additional funding will also support hiring across engineering, product and go-to-market functions as customer demand continues to grow.

A key part of Foxglove’s impact on the robotics ecosystem is its open-source logging format, MCAP, which the company introduced to standardize how robotics data is recorded and shared. MCAP has since become widely adopted across the industry and is integrated with major robotics frameworks, helping developers work with diverse data types in a unified environment. This standardization effort has positioned Foxglove at the center of modern robotics data infrastructure.

Foxglove was founded by Adrian Macneil and Roman Shtylman, who previously worked on large-scale robotics and autonomous systems. The founders identified a persistent gap in tooling for robotics teams, many of which were forced to build custom internal solutions to manage and analyze operational data. Foxglove was created to replace these fragmented approaches with a scalable, purpose-built platform that can be deployed in cloud, hybrid or on-premises environments.

The company’s tools are now used by tens of thousands of developers worldwide, ranging from early-stage startups to large enterprises deploying autonomous systems at scale. Customers rely on Foxglove to shorten debugging cycles, improve system reliability and accelerate the transition from prototype to production. As Physical AI applications expand into sectors such as logistics, manufacturing, agriculture, aerospace and defense, the need for robust data infrastructure has become increasingly critical.

Investor backing from Bessemer Venture Partners, Eclipse Ventures and Amplify Partners reflects a broader trend of capital flowing into enabling technologies for robotics rather than end-user applications alone. These investors view data infrastructure as essential to unlocking the next wave of innovation in autonomous systems, much as cloud platforms and developer tools fueled earlier software revolutions.

With its Series B funding secured, Foxglove is focused on cementing its role as a core platform for Physical AI development. The company aims to help robotics teams spend less time building internal tooling and more time solving domain-specific challenges, ultimately accelerating the safe and reliable deployment of autonomous systems in the real world.