Distributed Storage Startup Tigris Raises $25M to Challenge Big Cloud with AI-Optimized Data Networks

The rise of AI has dramatically increased demand for computing power, leading companies like CoreWeave, Together AI, and Lambda Labs to gain traction by offering distributed compute capacity. However, most organizations still rely on the big three cloud providers—AWS, Google Cloud, and Microsoft Azure—for data storage. These platforms were designed to keep data near their own compute resources, making them ill-suited for the decentralized, high-speed demands of modern AI workloads. Ovais Tariq, co-founder and CEO of Tigris Data, argues that storage must evolve alongside compute. “Modern AI workloads and AI infrastructure are choosing distributed computing instead of big cloud,” Tariq said. “We want to provide the same option for storage, because without storage, compute is nothing.” Tigris, founded by former Uber storage architects, is building a network of localized data storage centers tailored for AI. Its AI-native platform moves with compute, automatically replicates data to where GPUs are located, handles billions of small files, and delivers low-latency access for training, inference, and agentic AI tasks. To support this mission, Tigris recently raised $25 million in a Series A round led by Spark Capital, with participation from existing investors including Andreessen Horowitz. The startup is directly challenging the dominance of “Big Cloud,” as Tariq calls the established providers. He points out that AWS, Google Cloud, and Azure charge egress fees—often referred to as the “cloud tax”—when data is moved between clouds or downloaded for use elsewhere. These fees can become a major cost burden. Batuhan Taskaya, head of engineering at Fal.ai, one of Tigris’ customers, said egress charges once made up the bulk of his company’s cloud spending. Beyond cost, latency remains a critical issue. Centralized storage systems struggle to keep pace with distributed AI infrastructure, creating bottlenecks that slow down model training and real-time inference. Tigris addresses this by placing storage close to compute, reducing delays and improving performance. Tigris’ customers are primarily generative AI startups focused on image, video, and voice models—workloads that require fast access to massive, latency-sensitive datasets. “Imagine an AI agent processing local audio,” Tariq explained. “You want the compute and the storage to be as close as possible.” For regulated industries like finance and healthcare, data sovereignty is another key concern. Keeping data within specific regions helps meet compliance requirements. Additionally, companies are increasingly asserting control over their data, especially after Salesforce blocked AI rivals from accessing Slack data earlier this year. Tariq believes organizations now recognize that data is the fuel of AI and want to own and manage it themselves. “They don’t want someone else in control,” he said. With the new funding, Tigris plans to expand its infrastructure, growing its network of data centers. Since its founding in November 2021, the company has grown eightfold annually. It currently operates in Virginia, Chicago, and San Jose and aims to extend into Europe and Asia, with planned centers in London, Frankfurt, and Singapore.