How Tigris is Building the Storage Layer for the AI Era

Tigris, building a globally distributed object storage service that unlocks true choice for builders, led by Ovais Tariq, Himank Chaudhary, and Yevgeniy Firsov, has secured $25 million in Series A, led by Spark Capital, with continued support from Andreessen Horowitz.

Tigris offers a globally distributed storage service designed specifically for AI. It enables developers to store and process training data at scale, serve models with low latency worldwide, and move freely across clouds without costly egress fees. Unlike traditional cloud storage treated as a secondary service, Tigris provides a foundation optimized for AI workloads, eliminating vendor lock-in, hidden costs, and performance bottlenecks that can slow innovation.

Tigris provides AI teams with storage tailored for their workflows, enabling training at scale, borderless inference, safe agent collaboration, flexible compute, and faster iteration. The new funding accelerates the deployment of proprietary hardware in datacenters worldwide, allowing Tigris to operate independently rather than relying on major cloud providers.

Tigris powers some of the leading multimodal and media-generation AI platforms, including fal.ai for high-speed generative AI, KREA for creative media, and hedra.com for multimodal AI applications. It also underpins developer-first cloud providers like Fly.io and Beam, making it the storage foundation for the fastest-growing AI and cloud companies.

The cloud market is poised for disruption. AI developers are already moving toward specialized compute providers like CoreWeave and Together AI for best-of-breed solutions, and storage is next. Challenging Big Cloud requires owning the full stack, from hardware to software, so innovation can happen without compromise, offering builders a faster, fairer, and lock-in–free alternative. Tigris positions itself as that specialized storage layer, setting a new standard for quality and innovation.