Space raises USD $2.4 million for AI-native filesystem

Space has raised USD $2.4 million in a pre-seed round led by a16z Speedrun.

Golden Ventures, Northside Ventures and a group of angel investors also participated. The angels have backgrounds in software companies including Parsec, Sentry, Stan, Superwhisper and Modem.

The startup is building a distributed filesystem that lets users access cloud-hosted files without using local disk space. It is developing what it calls an AI-native filesystem that sits above the operating system and makes remote files appear to be stored locally. Applications can open files through the system without downloading a full copy to a device.

Its approach streams only the byte ranges requested by an application or AI agent when needed. This is designed to reduce duplicated files, syncing and separate integrations between storage systems and software tools.

Access layer

Space is targeting a problem that has become more visible as teams work with larger design files, software repositories, datasets and media libraries across multiple devices and cloud services. It argues that while cloud storage expanded capacity, it did not remove the need to move, copy or ingest data before work could begin.

The issue also affects AI tools, which often require users to upload or index files before an automated workflow can start. Space wants agents and people to work from the same live filesystem rather than through separate pipelines.

The system is designed to work with existing desktop software. Files appear in Finder and can be opened in native applications, including video editing tools, design software, code editors and CAD programs.

Space contrasts its model with storage products such as Dropbox, Box and Google Drive, which are mainly built around syncing files to local devices or shifting activity into browser-based workflows. Its product works at the filesystem level, allowing applications to use the same shared data layer without tool-specific connections.

Examples the company gives include a video editor opening a project larger than the local machine, an architect working with a large building model and a developer accessing a full code repository without storing it on a laptop. It also says AI agents could navigate large stores of organisational data and read only the portions relevant to a specific task.

Founding problem

Space was founded by Matthew Ao, Arihant Bapna and Jason Zhao. The idea came from their own experience handling large volumes of media and company data.

Zhao first encountered the issue while managing dozens of terabytes of video footage accumulated over years of work on YouTube. The problem later widened at the founders' previous company, where the team moved terabytes of footage each month and lost time to uploads, downloads and file compression.

The first prototype was built in late 2025. Space later based itself at Founders Inc. in San Francisco as it raised the pre-seed round and further developed the product.

Space is currently in private beta with about 100 users and teams onboarded. It also says it has built an audience of more than 80,000 across platforms.

Its initial focus is on teams in video, marketing, and architecture, engineering and construction, where local storage limits can become an obstacle. The company also sees those sectors as early adopters of workflows in which people and AI systems need to work on the same large sets of unstructured data.

Beyond that, it intends to expand into AI-focused companies and more data-heavy areas including media, training data infrastructure, computer vision and enterprise data systems.

Investors in the round framed the product as a challenge to a long-standing assumption in computing: that a file must first reside on a device before it can be used.

"Humans and agents are working across more data than any local disk can hold. They should not have to wait for files to be copied or ingested before work can begin," said Jason Zhao, co-founder of Space.

"The filesystem is the layer shared by every application, device, user, and agent. Rebuilding it around live, distributed data makes every workflow above it faster," said Arihant Bapna, co-founder of Space.

"The computer no longer needs to hold every file. It becomes a window into any file, wherever it lives and however large it is," said Matthew Ao, co-founder of Space.

Jonathan Lai of a16z described the company's technology as a new type of filesystem built for both people and software agents.

"Space is disrupting one of the oldest assumptions in computing: a file has to exist on your device before you work with it," said Jonathan Lai, general partner at a16z. "Matt, Ari, and Jason are innovating with a new type of AI-native filesystem that can serve both human creators and AI agents with the same primitive: instant access to the exact data they need."