Revolutionizing Data Storage: How This Startup Aims to Compete with Major Cloud Providers
The rapid rise of AI companies has created a seismic shift in the demand for computing power. As a result, innovative organizations like CoreWeave, Together AI, and Lambda Labs have stepped into the limelight, drawing significant investment and attention for their ability to deliver distributed compute capacity. However, many businesses still find themselves relying on the traditional giants of the cloud—AWS, Google Cloud, and Microsoft Azure—whose storage systems are primarily designed to keep data close to their own compute infrastructure, limiting scalability across diverse clouds or geographical regions.
Embracing Distributed Computing
“Modern AI workloads and infrastructure are increasingly opting for distributed computing rather than sticking with the big cloud providers,” explains Ovais Tariq, co-founder and CEO of Tigris Data. “Our goal is to extend this concept to storage. After all, without storage, compute is ineffective.”
Tigris’s Innovative Approach
Tigris was founded by the talented team responsible for creating Uber’s storage platform. The company is on a mission to establish a network of localized data storage centers tailored to modern AI requirements. Tariq described their AI-native storage platform as one that effortlessly “moves with your compute,” facilitating automatic data replication to where GPUs are located. The system supports billions of small files and ensures low-latency access essential for tasks like training, inference, and agentic workloads.
Recently, Tigris secured $25 million in Series A funding, with Spark Capital leading the round, alongside contributions from existing investors such as Andreessen Horowitz. The startup is poised to challenge the incumbents, which Tariq refers to as the “Big Cloud.”
Image Credits: Tigris Data
Rethinking Cloud Costs
Tariq highlights that these major cloud providers not only present a costlier data storage solution but also a less efficient one. For instance, many have historically imposed egress fees—often termed “cloud tax” within the industry—when customers wish to switch to another cloud provider or transfer their data for tasks like utilizing a more affordable GPU or training models across various locations. Imagine needing to pay extra just to stop going to your gym—that’s the reality for many businesses today.
According to Batuhan Taskaya, head of engineering at Fal.ai, a customer of Tigris, these egress fees once consumed a significant portion of their cloud expenditure.
Addressing Latency and Other Challenges
Beyond the financial burden of egress fees, Tariq points out an even more pressing issue: latency. “These fees are merely a symptom of a larger problem: centralized storage cannot keep pace with the decentralized, high-speed AI ecosystem,” he notes.
With over 4,000 customers, many of whom are generative AI startups developing models for images, videos, and voice, Tigris’s services are crucial. “Consider interacting with an AI agent managing local audio. You demand the lowest latency, which requires both your compute and storage to be in proximity,” Tariq explains.
The Case for Localized Storage
Large cloud providers often fall short when it comes to optimizing for AI workloads. The process of streaming massive datasets for training or running real-time inference across multiple regions often leads to latency bottlenecks, hampering model performance. However, having access to localized storage accelerates data retrieval, allowing developers to execute their AI workloads effectively and affordably using decentralized cloud solutions.
“Tigris enables us to scale our workloads in any cloud while providing access to the same data filesystem from various locations—without the charge of egress,” Taskaya states.
The Importance of Data Security and Ownership
In regulated industries like finance and healthcare, ensuring data security is paramount, often deterring organizations from implementing AI solutions. Tariq also observes the growing demand among companies to retain ownership of their data. This sentiment was clearly illustrated when Salesforce restricted its competitors from accessing Slack data earlier this year. “Companies are becoming increasingly cognizant of the significance of their data and how it fuels large language models and AI integration. They desire greater control,” Tariq explains.
Future Growth and Expansion
With fresh capital in hand, Tigris aims to continue developing its storage centers to meet rising demand. Since its inception in November 2021, Tariq proudly states that the startup is growing 8x annually. Currently, Tigris operates three data centers in Virginia, Chicago, and San Jose, with plans to expand further in the U.S. and internationally in locations such as London, Frankfurt, and Singapore.
As the AI landscape continues to evolve, staying attuned to advancements in data storage and computing solutions will be crucial for companies seeking to harness the full potential of this transformative technology. For individuals and businesses eager to remain competitive, exploring innovative options now can pave the way for a brighter, more efficient future.
Are you ready to explore how localized storage can revolutionize your operations? Join the conversation with Tigris and empower your AI ambitions today!
