Weka recently raised $140 million in Series E funding, with aim of addressing data challenges.
Even though more businesses are investing in AI, many are having difficulty implementing AI-powered initiatives, and even fewer are seeing a significant return on their investment.
There are numerous difficulties. However, data management is one that comes up frequently. Companies require segregated, unoptimized, and disorganized data in order to train, run, and fine-tune AI models. The open-source data benchmarking platform Great Expectations conducted a poll in 2022 and found that 77% of firms expressed concerns about the quality of their data.
Startups receiving a ton of financing are making promises to solve these data issues.
Weka, a platform for creating data pipelines that manage a variety of data sources, kinds, and sizes, revealed on Wednesday that it had raised $140 million in a two-part Series E round ($100 million and $40 million), with participation from Nvidia, Qualcomm Ventures, Hitachi Ventures, Micron Ventures, and other investors. Valor Equity Partners led the round. The oversubscribed round has now valued Weka at $1.6 billion post-money, quadrupling its previous valuation.
Liran Zvibel, the CEO of Weka, and Maor Ben-Dayan and Omri Palmon, two of the company’s other co-founders, got together while developing the data storage business XIV, which IBM purchased in 2007 for $350 million. After working for IBM for a while, the three eventually left to pursue their own independent endeavors.
However, Zvibel claims that the issue of data management continues to bother him.
He expressed his frustration and disillusionment at seeing clients compelled to adopt fragmented, isolated data infrastructure solutions that were expensive, time-consuming, and difficult to set up, run, and maintain. “With the emergence of cloud computing, high-performance computing, machine learning, and the first AI workloads, the issue became particularly evident.”
Zvibel hired Ben-Dayan and Palmon in 2013 to develop a new set of data tools that might enhance the movement, storage, and management of data.
According to Zvibel, “We envisioned a platform powerful enough to support large-scale, data-intensive workloads in demanding and distributed environments, as well as the performance demands of next-generation compute hardware.” “To meet the demands of modern workloads, we knew it would need to process tens of terabytes of data and be deployable anywhere.”
Weka’s main product is a parallel file system, which is a type of distributed file system that can distribute and coordinate data processes (like copying files) simultaneously across multiple locations (such as workstations and servers). Furthermore, Weka offers services and solutions to serve high-performance compute workloads in public clouds, hybrid clouds, on-premises data centers, and visual effects.
Zvibel asserts that Weka’s architecture has the ability to accelerate AI model training by cutting down on the time required to copy data between storage locations. He asserted that a typical generative AI data pipeline includes numerous steps of copying data sets, thereby wasting crucial training time. “Weka continuously feeds data into model training hardware to expedite the training process.”
Weka competes with data platforms such as DataDirect, Pure Storage, NetApp, and Vast Data. Having completed a $118 million Series E investment round in December 2023, which increased the startup’s valuation to $9.1 billion, Vast is one of the more formidable of the group.
With more than 300 businesses as clients, including the AI firm Stability AI, 11 Fortune 50 companies, and a number of unknown domestic and international government bodies, Weka seems to be holding its own, though.
Zvibel stated that despite having about 400 workers globally and intentions to add another 25% within the next year, the Silicon Valley-based company now has a “line of sight” to cash flow positivity by December 2024.
He continued, “Weka was able to raise at extremely favorable and advantageous terms because of the recent raise, which was calculated based on favorable market conditions and proactive investor interest.” We anticipate our average burn rate to be less than half a million per month before we reach that milestone. Our recurring revenue has surpassed $100 million annually, and we continue on our hyper-growth trajectory.