WEKA is a technology company that builds a high-performance, software-defined data platform optimized for AI, machine learning, and high-performance computing (HPC) workloads. Its core product, the WEKA Data Platform (now evolving into NeuralMesh™), enables storage, processing, and management of massive unstructured datasets across edge, core, cloud, and hybrid environments, delivering 10-100x performance gains while reducing energy use.[1][2][3]
WEKA serves leading AI innovators, research teams, and enterprises in sectors like autonomous driving, healthcare, finance, and manufacturing, solving critical bottlenecks such as data silos, scalability limits, and low GPU utilization in compute-intensive applications. The platform achieves up to 93% GPU utilization for AI model training and inference, with recent advancements like Augmented Memory Grid extending GPU memory 1000x and cutting time-to-first-token by 20x.[2][3]
Founded in 2013 in Campbell, California (formerly WekaIO), WEKA was started by a team of cofounders including Liran Zvibel (CEO), Amit Kanner, Omri Palmon, Maor BenDayan, and others, who spotted gaps in traditional storage for modern unstructured data and compute-heavy workloads long before AI's mainstream rise.[1][2] With expertise in storage systems and finance (e.g., Brian Goldfarb, an HBS alum), they aimed to disrupt the status quo by building a screaming-fast, software-defined parallel filesystem from scratch.[2]
Early traction came from evolving this into a scalable data platform for speed and simplicity across environments. Pivotal moments include pioneering containerized microservices with NeuralMesh™ for adaptive AI infrastructure and partnerships validating its "green data" efficiency, as noted by investors like Lumir, positioning WEKA as a leader in AI/HPC innovation.[2]
WEKA rides the explosive growth of agentic AI and reasoning models, where data infrastructure must match trillion-parameter scales and real-time demands amid surging unstructured data volumes.[2][3] Timing is ideal post-2023 AI boom, as traditional storage fails GPU clusters, while market forces like energy constraints and multi-cloud adoption favor WEKA's efficient, adaptive solutions over competitors like SandStone or Tintri.[1]
It influences the ecosystem by pioneering "augmented memory" and NeuralMesh, enabling breakthroughs in autonomous systems, healthcare AI, and HPC; this shifts paradigms toward sustainable, high-ROI infrastructure, accelerating innovation for hyperscalers and startups alike.[2][3]
WEKA is primed to dominate AI data infrastructure with NeuralMesh expansions, new WEKApod economics, and deeper integrations for trillion-scale reasoning. Trends like edge AI proliferation, stricter energy regs, and memory-bound models will amplify its edge, potentially evolving it into the de facto platform for "green" AI factories.[2][3]
As AI shifts from training to inference at scale, WEKA's humility-driven leadership could redefine storage economics, turning data bottlenecks into advantages and solidifying its catalyst role in the reasoning era—echoing its founding mission to challenge the status quo.[2]
WEKA has raised $140.0M in total across 1 funding round.
WEKA's investors include 10D, Atreides Management, Bessemer Venture Partners, Lightspeed Venture Partners, Pitango Venture Capital.
WEKA has raised $140.0M across 1 funding round. Most recently, it raised $140.0M Series D in November 2022.
| Date | Round | Lead Investors | Other Investors |
|---|---|---|---|
| Nov 1, 2022 | $140.0M Series D | 10D, Atreides Management, Bessemer Venture Partners, Lightspeed Venture Partners, Pitango Venture Capital |
