DEVOPSdigest

RapidMiner announced the release of its platform enhancement, RapidMiner 9.6. This update prioritizes people – not technology – at the center of the enterprise AI journey, providing new, unique experiences to empower users of varying backgrounds and abilities.

The RapidMiner 9.6 platform embraces a broad array of users, enabling collaboration between coders, non-coding data scientists, and business users on a platform seamlessly combining automated data science, visual workflows, and coding as needed or preferred.

In enterprise analytics, there will always be different groups who bring different and unique skills, domain knowledge and experience to a project, but prefer to work with very specific toolsets – such as coding notebooks or BI platforms. RapidMiner 9.6 allows diverse teams to work together on the same project within their preferred experience paradigms.

This represents a foundational step in the company’s major initiative to drive easy collaboration between advanced data scientists, budding citizen data scientists and business leaders. The ultimate goal for RapidMiner is to get more models into production where they can deliver profound business impact – this is all part of an initiative to help prevent the spread of the Model Impact Epidemic, an ongoing issue in the world of data science that is wasting time, money, and effort.

RapidMiner 9.6 includes some major advancements:

- RapidMiner Go, a highly differentiated automated machine learning (ML) solution that’s built specifically for business users with no previous data science experience to evaluate the ROI of their data science models. It’s delivered through a browser, in SaaS and private hosted options, so that resources aren’t locally consumed on the users’ machines.

- JupyterHub is now built directly into the RapidMiner platform. Users can code in Python directly in a centrally managed and governed location. Diverse teams can then use code-based work interoperably with visual workflow-based projects and automated data science.

- RapidMiner Model Ops, part of the 9.4 release, can now deploy, manage, and monitor models that are custom-created – even those that are built entirely in Python.

- Integration with Grafana, the leading open source data visualization tool. Grafana can be used by anyone to quickly build interactive web apps and dashboards to share the results and insights that are generated by their models. This helps drive models successfully into production and share the impact of a data science initiative across an enterprise.

“Our goal with RapidMiner 9.6 is to help expand access to data science for users of every skill level,” said Dr. Ingo Mierswa, Founder of RapidMiner. “With this top of mind, the platform enhancement provides even more depth for experienced data scientists, Python coders and more, while simplifying the process for all in order to provide greater clarity for stakeholders and decision makers at each stage of the business process. Data Science is a team sport and RapidMiner 9.6 and the new RapidMiner Go are the best tools to support all players on the team."

RapidMiner Go is significant because it helps anyone frame business problems as machine learning problems, without any previous experience in ML. This means that non-data scientists who are close to the problem and the data can prototype a solution on their own and RapidMiner Go even helps to optimize their models for business profits. RapidMiner Go hooks directly into the rest of the RapidMiner platform so that models created by business users are easily tuned by more seasoned professionals. Models can also be “auto-deployed” instantly – with or without development operations (DevOps) teams. At the end of the process, it delivers a production-ready AI package including the fully tuned and immediately usable model, as well as supporting business case materials and a reusable, editable data pipeline.