Salesforce Open-Sources Automated Machine-Learning Library Used To Build Einstein

Salesforce Thursday released an open-source version of the automated machine-learning library it developed to power its Einstein artificial intelligence platform.

TransmogrifAI bundles capabilities that automate the development of machine-learning models by feeding structured data to pre-packaged algorithms.

"TransmogrifAI has been transformational for us, enabling our data scientists to deploy thousands of models in production with minimal hand-tuning and reducing the average turnaround time for training a performant model from weeks to just a couple of hours," said Shubha Navar, senior director of data science for Einstein, in a blog.

[Related: Salesforce Beefs Up AI Capabilities With Datorama Buy]

Salesforce recognized the end-to-end capabilities of its library could be useful to other businesses hoping to take advantage of AI, but struggling with a shortage of qualified professionals available for hire.

"We believe that every business today has more machine-learning use cases than it has data scientists, and automation is key to bringing the power of machine learning within reach," Navar said.

TransfmogrifAI is built on top of the Apache Spark big data framework. Salesforce, a contributor to that project, will continue developing the machine-learning library with the Spark community.

Salesforce started working on the technology three years ago as part of the project that yielded Einstein.

TransmogrifAI automates the machine-learning workflow with capabilities like automatically transforming data into formats algorithms can process.

Bluewolf, an early Salesforce partner that's owned by IBM, has studied the adoption of AI in the marketplace.

In a recent survey, the San Francisco-based systems integrator, which builds unique AI services around Salesforce and IBM technology, highlighted an "AI Investment Gap"—a disparity between C-level executives who understand AI and those who have successfully deployed it in their businesses.

That Bluewolf report noted 72 percent of respondents plan to deploy AI within the next 12 months.

"Salesforce’s move to open-source their automated machine-learning libraries is positive for the market, making powerful AI capabilities available to a wider audience of users, quickening the mainstreaming of AI adoption in core business applications, and improving the customer experience with AI," said Steve Faris, Bluewolf's chief platform officer.

Navar, in the Salesforce blog, reiterated Salesforce's commitment to open-source development.

"Machine learning has the potential to transform how businesses operate, and we believe that barriers to adoption can only be lowered through an open exchange of ideas and code," she said.

"By working in the open we can bring together diverse perspectives to continue to push the technology forward and make it accessible to everyone."