NEW YORK -- In a bid to enhance its AI-powered capabilities, expand available data sources for its business intelligence software and widen users' AI options, multinational software giant Oracle acquired AI sales enablement services company DataFox.
The Oracle-DataFox acquisition, completed Oct. 31, will likely give Oracle users access to DataFox's cloud-based AI data engine -- and to the applications the vendor has built with it -- once the software is incorporated into Oracle systems. Oracle also will continue to sell the DataFox AI engine as a stand-alone product.
The acquisition made sense for both companies, according to Bastiaan Janmaat, CEO and co-founder of DataFox, based in San Francisco.
"What we were building was just very aligned with the Oracle roadmap," he told us in an interview with him and Melissa Boxer, vice president of Adaptive Intelligent Applications at Oracle, during the AI Summit.
Oracle, for its part, has been looking to expand its AI efforts.
Fire up that data engine
DataFox has a data engine that can "suck in unstructured content ... and then pull out of that structured data about companies," Janmaat explained. "Well, right now, it's data about companies. But, long term, it could be about anything."
That technology has enabled DataFox to create customer relationship management software that can automatically find reference and signal data about companies and distill it down to relevant, focused bits of information -- "smart talking points," as Janmaat put it -- that can be used to provide sales reps with quick, useful information.
DataFox also sells tools for sales organizations that can automatically "surface companies that are a good fit based on past clients," Janmaat said. The software also automatically prioritizes based on a customizable weighting system.
The AI tools aren't meant to take jobs away from sales reps, Janmaat said, but instead augment their own abilities by automatically providing them with fast access to information about potential clients.
Humans plus AI
Bastiaan JanmaatCEO and co-founder, DataFox
Behind the scenes, the software -- still publicly available after the Oracle-DataFox acquisition -- uses AI-powered tools like machine learning algorithms and natural language processing (NLP), as well as continued human input from a dedicated team from DataFox that helps customize and train the AI-based models for high accuracy.
The DataFox team tracks accuracy, which might be in the 70s percentiles at first, and the team keeps tweaking the models "until it gets to high 90s," Janmaat said.
The DataFox engineers have strong domain authority, as well as machine learning expertise, Janmaat said.
They need to have that expertise, as the data engine draws in content from numerous places, including news sources, reference documents and company websites, which might not always have correct information, he said.
Companies rise, fall and change goals, so information must be constantly updated. Besides the human element, the Oracle-DataFox software also uses a blacklist to automatically block sources known to peddle fake news, like the satirical news site The Onion, for example.
The Oracle perspective
"We have a lot of great plans in place" for the data engine, Boxer said.
While saying she couldn't talk about the details of the Oracle-DataFox acquisition or a potential integration timeline due to a financial quiet period, Boxer noted that Oracle will start to use data from DataFox immediately, and it will integrate the AI components with Oracle's own applications later.
"We wanted the engine, because that will help us grow data and extend data to other domains," she said.
"This is a very differentiated position that they are taking, which is to say we are going to take care of the third-party data for you," Janmaat said, referring to the Oracle-DataFox acquisition.
The vendors said they don't expect the acquisition to disrupt the ability of DataFox users to integrate the software into other platforms through DataFox's API.
The AI Summit is Dec. 5 to Dec. 6 at the Javits Center.