Databricks Inc. this week completed its third acquisition in seven months aimed at improving its generative AI capabilities, purchasing Einblick Analytics Inc. The startup's natural language processing capabilities enable users of varied skill levels to work with data.
Financial terms of the transaction, which Databricks unveiled in a blog post Tuesday, were not disclosed.
In June 2023, Databricks reached an agreement to acquire MosaicML for $1.3 billion. The purchase added a platform that enables users to train and secure generative AI models using their proprietary data so that, unlike models trained exclusively on public data, the models can inform business decisions.
Four months later, in October 2023, Databricks agreed to acquire Arcion for $100 million. The vendor's tools added data ingestion and data replication capabilities that can be used to develop pipelines that feed and train generative AI models and applications.
Following those acquisitions, Databricks unveiled the Data Intelligence Platform, the next evolution of the lakehouse pioneer's Data Lakehouse Platform now in preview.
The new platform, including its rebranding, reflects the vendor's focus on making generative AI as essential to its portfolio as its longstanding data management capabilities. The Data Intelligence Platform, once finished, will join data platforms such as data lakehouses with analytics and data science.
Now, further demonstrating that focus on generative AI, Databricks is adding Einblick's natural language processing (NLP) capabilities.
Additive capabilities
Founded by researchers from MIT and Brown University, Einblick -- a German word that means insight -- is a startup based in Cambridge, Mass. It emerged from stealth in 2020.
Its tools translate natural language to code so that business users can create and run AI and machine learning models that would otherwise require the skills of a data scientist to develop.
Donald Farmer, founder and principal at TreeHive Strategy, called Einblick a strong fit for Databricks, given its potential to help more employees within organizations use analytics as part of their job.
For decades, analytics use within enterprises has stagnated. Studies vary, but for about 20 years, they have shown that only a quarter to a third of people within organizations use analytics as part of their job. The primary reason for the stagnation has been the complexity of analytics tools, with most tasks requiring coding skills.
Natural language processing, including generative AI chatbots and intuitive NLP tools such as Einblick's platform, has the potential to change that.
"Einblick's visual, collaborative data analytics capabilities seem like a strong fit and addition to Databricks' data and AI platform," Farmer said. "Einblick brings natural language interfaces and simple data workflow creation that could help bring Databricks' data intelligence in reach of business users."
Doug Henschen, an analyst at Constellation Research, likewise said Einblick is a good fit for Databricks. He noted that Einblick will help Databricks build its Data Intelligence Platform by providing natural language interfaces that allow non-experts to work with data while enabling experts to work more efficiently.
Einblick's tech will be used to enable natural language interaction directly with Databricks [as Databricks] brings together the formerly separate world of analytical data platforms … with analytics, business intelligence and data science capabilities.
Doug HenschenAnalyst, Constellation Research
"The Einblick acquisition is an interesting and good fit," he said. "Einblick's tech will be used to enable natural language interaction directly with Databricks [as Databricks] brings together the formerly separate world of analytical data platforms … with analytics, business intelligence and data science capabilities."
Einblick is not alone among analytics vendors in developing NLP capabilities. Like Einblick, ThoughtSpot developed its entire platform around natural language search and query capabilities. Meanwhile, vendors including Tableau and Qlik added NLP tools to their previously existing tools.
What differentiates Einblick from other analytics vendors, however, is that its NLP capabilities don't simply enable exploration and analysis.
Enhanced over the past year through integrations with emerging generative AI capabilities that expand the vocabulary of the vendor's capabilities, they also make data science a focal point, Henschen noted.
"What Einblick offers is a notebook-like analytics and data science environment that was enhanced with emerging generative AI capabilities," Henschen said. "This combination of data platform capabilities with analytics and data science capabilities is a developing trend [with] GenAI … a key catalyst of the trend."
Farmer likewise noted that more than merely NLP, Databricks' acquisition of Einblick is significant due to Einblick's use of NLP to enable data science.
"Einblick adds capabilities to translate natural language questions into executable data workflows using [coding languages] SQL and Python, machine learning models and more," he said. "That bridges the gulf between the capabilities of business users to describe what they need and Databricks' existing [developer-oriented] capabilities."
Following the acquisition, Einblick's leadership team including co-founder and CEO Emanuel Zgraggen and fellow co-founder and chief experience officer Philipp Eichmann are now part of Databricks.
The addition of talent may be just as significant as the addition of new technology, according to Farmer.
"Talent is tough to hire these days," he said. "They are getting some folks with a proven track record as a coherent team. That's a win."
How big a win depends largely on Databricks' plans for Einblick, Farmer continued.
If Databricks further develops Einblick into a high-level tool and extensively integrates its capabilities, the addition of talent and technology will be important. But if Einblick and its leadership are tucked into the existing Databricks portfolio without much integration, the talent and technology additions will be less meaningful.
Henschen, meanwhile, noted that like Sisu, Einblick was not able to gain significant market share as an independent vendor. That perhaps suggests that Einblick's technology is not so differentiated that the acquisition isn't equally about adding talent as adding new capabilities.
"Much like Sisu, Einblick wasn't exactly setting the world on fire, so it's easy to speculate that both acquisitions were as much about talent acquisition, market signaling and planned technology experimentation," Henschen said.
Seven benefits of generative AI for the enterprise.
Next steps
Following Databricks' acquisitions of MosaicML, Arcion and now Einblick, Henschen said he'd like to see Databricks Data Intelligence Platform begin to materialize.
The vendor has unveiled its plans to develop the portfolio, combining data platform capabilities with advanced analytics and AI, but it has not produced it.
"The Data Intelligence Platform is a very interesting and promising-sounding development," Henschen said. "But I'd really like to see concrete examples of what it can do to democratize insight through natural language interaction and/or deliver more and better-integrated capabilities at a lower cost than currently achieved through the use of separate products."
Similarly, Farmer said that it's time for Databricks to concentrate on building the Data Intelligence Platform it has promised and use the various platforms it has acquired over the past seven months to do so.
"The next step [is to] integrate, integrate, integrate," he said. "And [Databricks should] use these natural language capabilities to reimagine the Databricks workflow for business users, not just tacking on another feature. So [it should] design, design, design, too."
Eric Avidon is a senior news writer for TechTarget Editorial and a journalist with more than 25 years of experience. He covers analytics and data management.
