Sergey Nivens - Fotolia

DataRobot training aims to upskill citizen data scientists

DataRobot unveiled a new data science training program, and enhancements to its enterprise AI platform, including a geospatial capability and automated time series anomaly detection.

A new training program from AI and auto machine learning vendor DataRobot aims to teach citizen data scientists, including business analysts and data analysts, practical data science and AI skills.

The paid program, "10x: The Applied Data Science Academy," provides instruction in skills such as problem framing, exploring data, feature engineering and deploying models. It involves 40 hours of hands-on, self-paced training, 20 hours of practical labs and 40 hours on a capstone project.

With the program, revealed on Wednesday during DataRobot's AI Experience Worldwide virtual conference, held June 16-17, DataRobot enters an already crowded online AI training field. Vendors such as Google, IBM and Microsoft have long offered free and paid analytics and AI training programs, as have many colleges and universities.

Like DataRobot's 10x Academy, many such training courses are available directly through vendors' websites. Others are offered on online learning platforms including EdX and Udemy.

"There are now many of these programs and certifications," said Dave Schubmehl, research director for cognitive/artificial intelligent systems and content analytics at IDC.

"Some of these training programs that provide skills and education in areas such as developing deep learning models can be very useful and appeal to a broad set of employers," Schubmehl continued. "Others that focus on a particular product or tool set may not be as useful, especially if those tools aren't widely used."

DataRobot training, DataRobot
DataRobot unveiled a new training program and updates to its enterprise AI platform

The new program expands on DataRobot's other training offerings. DataRobot University provides courses for data analysts, data scientists and educators. The Boston-based vendor's Academic Support Program matches DataRobot with colleges and universities for joint programs.

While those programs may offer similar skills, "the 10x Academy focuses on empowering individuals through relentlessly practical hands-on training to develop and increase their skills, rather than partnerships with companies or universities," Dan Wright, COO of DataRobot, said in an email.

There are now many of these programs and certifications.
Dave SchubmehlResearch director for cognitive/artificial intelligent systems and content analytics, IDC

The 10x Academy is principally designed for citizen data scientists, although technically no data science experience is required, and provides those who complete it with a standardized applied data science professional certification from DataRobot.

"Now more than ever, organizations have a massive need for data scientists who have the practical experience, knowledge and skills needed to solve complex business problems and create business value using AI," Wright said. "We believe our program will help individuals develop these much-needed skills and provide an accessible, accelerated path to a career in this field."

The program will also include a trial version of DataRobot's enterprise AI platform, access to the DataRobot community, mentoring and job placement resources. Wright declined to say how much the program costs but noted that it's free for the first 100 participants who finish it.

In related news, DataRobot also unveiled new enhancements to its enterprise AI platform.

Bringing the platform to version 6.1, the enhancements include automated time series anomaly detection within the Automated Time Series product, a feature to test production models against other models to see how well they compete over time in DataRobot MLOps (machine learning operations), and a hub to collaborate with team members on AI initiatives, enabling users to manage and organize machine learning project assets around use cases.

The update also provides the ability for users to set specific conditions to trigger when a model does not have confidence in a prediction, as well for a feature for users to add geospatial data to their predictive models.

Geospatial data in AI applications can be quite important, Schubmehl said of the new capability.

Using geospatial data, for example, a user could build a targeting application based on who might go to certain restaurants. The application could take in data on where prospective diners are located in relation to the restaurant or at what times of the day they tend to go there, Schubmehl said.

With that data, an AI model could then send automated offers to prospective customers. Such offers could be more personalized and possibly more effective than conventional marketing.

Other applications could include shipping and logistics, traffic management and store placement, he added.

Next Steps

New DataRobot CEO sees bright AI future for the vendor

DataRobot looks to enterprise AI coders with Zepl acquisition

Dig Deeper on AI technologies

Business Analytics
Data Management