WEKA operates in the data infrastructure sector, with a focus on high-performance storage. The American company sells software capable of pooling resources from multiple servers into a distributed file system. Its goal: to enable applications to access data quickly, both on customers’ on-premises infrastructure and in the cloud. It serves organizations that use computing clusters and graphics processing units, among others.
A company born out of the limitations of traditional storage
Founded in 2013, WEKA has Israeli technological roots and is headquartered in California. Liran Zvibel, co-founder and chief executive, previously worked on storage systems at XIV, a company acquired by IBM. The project stems from an observation: legacy architectures do not always meet the needs of parallel applications, which generate numerous simultaneous accesses to files of widely varying sizes.
WEKA therefore developed a software architecture designed to leverage flash memory and fast networks. The company has gradually linked its development to that of artificial intelligence. In 2024, it announced a funding round led by Valor Equity Partners, intended to support its expansion and the development of its platform.
Keeping data flowing to applications
The WEKA platform relies on a parallel file system that distributes data and operations across multiple machines. In particular, it uses NVMe drives, a flash storage technology suited to fast access. This approach aims to combine high throughput, low latency and the management of large numbers of files, without requiring applications to completely rewrite the way they access data.
The software can be deployed on compatible servers or in public cloud environments. It also allows a fast storage tier to be combined with object storage, so that data placement can be adapted to performance and capacity requirements. WEKA therefore does not replace a database: it provides an infrastructure that underpins processing tasks, files and production pipelines.
Target use cases include AI model training, scientific computing, life sciences research and media. For these customers, a key concern is preventing expensive computing resources from sitting idle because they are not receiving data quickly enough.
What next?
The proliferation of AI projects opens up a market for WEKA, but also intensifies competition among storage specialists, major hardware vendors and cloud providers. The company will need to demonstrate its benefits on real-world workloads, beyond performance benchmarks alone. Cost control, ease of operation and integration with existing infrastructure will matter just as much as speed. Its growth will also depend on its ability to support the transition from AI experiments to services running sustainably in production.