your123 -

DataRobot platform gets new tools for data scientists

The AI vendor released Core and 7.3. Core uses Zepl's technology, while 7.3 enables data science teams to employ models that combine complex rules and business logic.

DataRobot today introduced two new offerings for data science experts: Core and DataRobot 7.3.

DataRobot Core widens the vendor's AI Cloud platform to accommodate code-first data scientists and expert business users. It extends the platform's capabilities with enhanced decision intelligence.

The new system widens the market for the vendor, which built its business model by appealing to non-data scientists, said Mike Gualtieri, an analyst at Forrester.

Filling a gap with Core

The significance is DataRobot is trying to serve multiple personas with their platform.
Mike GualtieriAnalyst, Forrester

"The significance is DataRobot is trying to serve multiple personas with their platform," Gualtieri said. "It fills the coding gap that they had."

With Core, DataRobot is trying to serve larger enterprises with multiple needs and different types of talents -- coders and knowledge workers who don't have that skill and knowledge, Gualtieri said.

Before this offering, DataRobot's platform had fallen behind others -- such as IBM's -- that serve varied work personas. The competitive platforms offer a  comprehensive experience for professional data scientists, citizen data scientists, application developers, ITOps workers and anyone else involved in AI projects.

Gualtieri said that with Core, DataRobot is giving developers working with code access to data preparation tools and machine learning capabilities.

"By being in what they're calling the Core, now these data scientists have access and can benefit from all of the other capabilities that … simply don't exist in a coding environment," Gualtieri added.

Making use of their acquisitions

With the new system, the vendor is using technology it gained from it gained from Zepl, the cloud data science and analytics vendor it acquired earlier this year.

Core is built on Apache Zeppelin, an open source project. However, Gualtieri said that DataRobot's reliance on Zeppelin versus the more popular open source Jupyter notebook may give some data science teams pause.

"Most people know Jupyter notebook, they want to use Jupyter notebook," he continued. "There's technical reasons why Apache Zeppelin is probably better than Jupyter, but those aren't widely understood."

According to DataRobot, Core enables data scientists to move between code-first and automated model generation. Data scientists get access to open API that gives them programmatic access to the AI Cloud platform.

The significance of 7.3

Meanwhile, DataRobot also unveiled the latest update of its main platform, DataRobot 7.3, with 80 new features aimed at enabling AI-driven decisions across business lines.

DataRobot 7.3 enables data science teams to run anomaly detection with images and use the next generation of DataRobot's Text AI system and tools such as multimodal clustering, time series segmented modeling and multilabel classification. It also enables teams to deploy models that combine complex rules and business logic.

Gualtieri said the features are interesting because they enable enterprises to combine business rules with models.

"Business rules are human decision logic," he said. "A machine learning model is based upon the analysis of data."

Both DataRobot Core and DataRobot 7.3 are generally available now.

Dig Deeper on Machine learning platforms

Business Analytics
Data Management