Rubrik is creating another pivot to create a data lake of generated AI through new acquisitions.
SaaS data backup platform vendor Rubrik announced last week that it had agreed to acquire AI specialist Pretabase. According to a Rubrik spokesperson, the acquisition will bring Genai capabilities to data stored within Rubrik's platform.
Successful AI applications in the enterprise rely on large data stores, says Christophe Bertrand, an analyst at TheCube Research. Since backup platforms already have data from many companies, bringing that data closer to AI can drive projects.
“They really position themselves as the source of their data,” Bertrand said. “This acquisition is about accelerating the ROI of our clients' AI.”
Rubrik declined to answer questions from Informa TechTarget regarding acquisition costs, operations and other details. Other media such as CNBC reported that Rubrik intends to pay Predibase between $100 million and $500 million, saying that after the transaction is closed, the company will operate as a separate unit.
Next Pivot
Predibase sells a tool for tweaking what is called a small language model (SLM). These are derivatives of large-scale language models (LLMs) such as the ubiquitous ChatGPT and Google Gemini for their specific uses.
The vendor's platform allows customers to refine open source models of Meta's Llama and Mistral AI to customer SLM for specific workloads and applications, according to Predibase.
SLMS is likely the future of AI within many companies, says Scott Hebner, an analyst at TheCube Research. LLMS introduces external knowledge and security risks, complicating AI deployments compared to a more sophisticated knowledge base, and solves the solutions SLM can offer.
“There's no longer a time for businesses using common and publicly available AI,” Hebner said.
In a blog post detailing the acquisition, Bipul Sinha, CEO of Rubrik, wrote that the Rubrik platform created the data lake. With Rubrik supporting the creation of Data Lake, the vendor recently announced another industry pivot after dipping its toe into cybersecurity.
They are still a backup and recovery company. Don't make yourself a child.
Christophe BertlandAnalyst, TheCube Research
Rubrik is very different to implementing Data Lake features, he said, which helps create data lakes. Such motivation requires vendors to establish partnerships and connectivity.
“They are still a backup and recovery company. Let's not make ourselves children,” Bertrand said. “[But] How do you do it? [AI] When you don't come from that space? ”
Rubrik was the first backup vendor to explicitly discuss data using it within backups of LLM or SLM implementations, said Jerome Wendt, CEO and Founder of Data Center Intelligence Group.
Cohesity's Gaia Generated AI service can query backups within the platform, but is currently not exporting data from other applications.
According to Went, another major challenge facing Lubric is the clients themselves. While companies have been reviewing and analyzing AI for the past few years, many have found that there is a lack of accurate and relevant data accurate to build AI data lakes and agent workflows.
“The biggest problem for an organization is being the quality of the data,” Went said. “Many organizations don't handle analogies or petabytes. [of files]however, they still need to look up that data and ingest it. ”
Tim McCarthy is a news writer for Informa TechTarget, which covers cloud and data storage.
