26 February 2020


Putting People at the Center of AI: RapidMiner 9.6

There have been some major advancements to the RapidMiner platform since this article was originally published. We’re on a mission to make machine learning more accessible to anyone. For more details, check out our latest release.

Our goal here at RapidMiner is to make AI available for anyone—anyone who wants to use machine learning and artificial intelligence to positively shape the future. So how do we build tools that are accessible to anyone, no matter who they are or how they work?

That question has helped orient the work that we’ve been doing over the past several months, and which we’re excited to announce today as RapidMiner 9.6. The latest version of our platform puts people at the center of the AI journey. It does this by delivering unique experiences tailored to each team members’ unique skillsets and preferences, so that anyone can be involved.

What’s new with 9.6?

Details about the features and integrations that we’re rolling out are further down this post, but here’s a TL;DR if you want to get a high-level picture of what’s new:

With these improvements, we’re getting closer than ever to a world where anyone who wants to harness the power of AI and ML has the tools they need at their fingertips. Let’s jump into the details!

9.6 is made for speed — RapidMiner Go

RapidMiner Go is the intelligence of RapidMiner Auto Model, in a presentation layer built for business users, subject matter experts and business analysts. RapidMiner Go’s straightforward UI guides you through the process of turning a business challenge into a machine learning model. Available as part of AI Hub, RapidMiner Go shows you the data & models that will be the most effective at driving business impact and even allows you to project profits that will be derived from putting the model into production.

RapidMiner Go provides the business users an explainable ML package that includes a usable model, supporting business case materials, and a reusable and editable data pipeline. This whole process only takes a few minutes. You can even instantly “auto-deploy” your new model—with or without data science or DevOps support—as a web service. Or simply score a new data set with your brand-new model, to further evaluate the impact, or support ad-hoc decisions.

What’s especially powerful about RapidMiner Go’s automated workflow is that, in addition to delivering insights for users without access to a dedicated data science team, it also helps those who are working with data scientists. For example, data scientists can use models from RapidMiner Go as prototypes and either tune them further or use the output for their own work. This embraces the ‘data science as a team sport’ approach that many of our customers strive for. The approach makes everyone more productive – especially the data scientists.

But that’s not all we have for data scientists in 9.6. In addition to making ML more accessible for novices, we also wanted to build out our capabilities for the other end of the spectrum—those with comprehensive data science and coding experience. We’ve done this by creating a deep connection between RapidMiner and the Python coder’s favorite tool, Jupyter Notebook.

9.6 speaks Python — JupyterHub is built-in

By incorporating notebook environments into RapidMiner, we’ve made it easier for code savvy team members to work with RapidMiner and easily package the work that they do for others to use. We’ve made this possible by integrating a notebook environment into RapidMiner and through the use of Docker, co-deploying of JupyterHub with RapidMiner AI Hub (formerly RapidMiner Server). The tight integration even includes a single sign-on experience and an easy connection to the RapidMiner AI Hub repository.

This seamless linking of Python code and RapidMiner workflows makes it easy for coders to ensure that their coding efforts are re-used across the organization, they can collaborate more efficiently with non-coders, and they can even re-use other people’s work more effectively to enhance their coding efforts where appropriate (for example, a model created by someone without ML experience using RapidMiner Go).

9.6 is visual — easy visual insight

A model without actionable insights provides no benefit for an organization. As we discovered during our research into the Model Impact Disaster, less than 1% of the models that could have business impact are being utilized effectively. To help address this problem, we’ve baked easy-to-use dashboarding capabilities into the latest RapidMiner release.

This integration lets you take the insights that you’ve derived from RapidMiner and create real-time, interactive predictive and prescriptive dashboards that are consumable by the whole organization. This helps you more effectively communication the value and impact of the models that you’re using, which helps tremendously with the operationalization of the models that you’re building. People are inherently skeptical of ML/AI – both at the leadership level and the ‘boots on the ground’ level. A picture is worth a thousand words in demonstrating the potential for your models to make an impact and our new easy-to-use visualization tool offers an extremely broad array of charts and graphs to help tell your story and share your results.

If you’re still thinking about ways that AI and ML might be able to improve your business, but not sure where to start, you can sign up for a free AI Assessment. We’ll walk you through various use cases unique to your business and how they can improve your bottom line.

Related Resources