Design + Machine Learning: How IBM Designed an Award-Winning Data Platform

We recently celebrated a big moment in IBM’s design history. In recognition of its design for IBM Data Science Experience, IBM Design in San Francisco has been awarded this year’s Red Dot Award for Communication Design.

The Red Dot Award is of huge significance for IBM. It’s a formal recognition of the investment in design, design thinking and IBM Design, from what’s widely regarded as the most prestigious design award in the industry — a visible fruit of design labor towards driving meaningful outcomes.

The Red Dot Award is even more meaningful because the team of designers behind it were striking out into terra incognita: the field of data science was both nascent and tremendously complex, and machine learning, one of the most important integrations in the platform, was moving so quickly the team had to get used to the feeling of designing for rushing water.

So, how did IBM build an award-winning platform for data science? Here’s how our designers made it happen for the Data Science Experience — complete with insights and stories from the team that brought it to life.

In the Beginning

The design team, led by Director of Design David Townsend, started with a simple goal: to develop a one-stop data science workspace that would include all the collaboration and open-source tools data scientists need and use every day.

We needed to create a central community for Data Scientists to go to, a place they would collaborate and share in an open source environment. — David Townsend, Director of Design, IBM Analytics

The platform is a machine learning collaboration space, an all-in-one data environment where data scientists can analyze nuanced datasets, visualize insights, share their work with teammates and create, train and deploy machine learning models on a single platform.

Building a product that unifies these data processes requires straightforward, end-to-end design that’s enjoyable for all involved — the ultimate marriage of data and design.

And that’s what Data Science Experience aims to achieve.

Research, research and more research

The design process started with user research — understanding data scientists and mapping their needs, goals and day-to-day workflows.

Here’s what they found: data scientists’ workflows were fragmented — they had to toggle back and forth between a variety of workspaces and tools to get the job done. They would use Data Shaper to clean data, Jupyter for modeling and MatPlotLib for visualization. These tools support a linear process, but data scientists’ workflows are more cyclical — like this:

The vision for Data Science Experience was that it would accommodate this agile workflow, simplify the experience of working with data, and bring all the tools into a unified data ecosystem.

Data Science Experience is a beautiful manifestation of the power of user research to understand our users’ needs, challenges and motivations. — Caroline Law, Design Lead

The design team identified some of data scientists’ biggest needs:

· To work with fellow data scientists and learn from each other
 · To share algorithms and exchange data analysis techniques
 · To publish the results of their work and collaborate with peers across neighboring disciplines: data engineers who can help them prepare data and business analysts who can translate their insights into data-informed decisions

Data Science Experience can deliver on each front while keeping the experience simple and straightforward.

Designing a Solution

I found inspiration for Data Science Experience from a website my son uses for writing poetry. It is a community built to help writers become better poets through sharing and feedback, converged with very basic text editing tools. We were attempting to do the same thing for data scientists, but of course the suite of creative tools in Data Science Experience is many orders of magnitude more sophisticated.- David Schultz, Head of Studios, IBM Analytics, San Francisco

The team wanted the experience to be easy to use and accessible for companion sites built for data engineers, system administrators and other user personas. They started work on building the ultimate data ecosystem, an environment that would intuitively connect related data functions and allow easy collaboration.

Our team has shipped a product that will change the industry for many years — I’m proud of that. -Jason Azares, UX Design

In the process of designing Data Science Experience, the San Francisco team also developed interface frameworks that are now used and applied across IBM.

We built a UI framework that simply and clearly localizes core interactions into repeatable patterns that work for many different tools. This framework is now being used by several design teams to build multiple sites, and we’ve been very gratified to see that it works beautifully. — Valeria Montrucchio, Design Lead

Red Dot: A Win for the Future

IBM Data Science Experience is a leading offering in the market and this award showcases the design thought, rigor and intellect behind the ground breaking product design. The Data Science design team should be immensely proud of bringing in divergent thinking and innovation into the very complex space of data science.

It’s been cool to see how easily the framework of DSX has expanded to include Machine Learning, new open-source tools, and additional capabilities from IBM. — David Schultz

Gartner’s Magic Quadrant for Data Science Platforms shows IBM as a leader. The report states : DSX is likely to be one of the most attractive platforms in the future — modern, open, flexible and suitable for a range of users, from expert data scientists to business people¹

Data science is complex enough without clunky tools and processes getting in the way. Data Science Experience was designed to lend clarity and uniformity to otherwise disparate data processes and become a user-friendly tool in a data scientist’s arsenal — one that continues to grow, incorporating deeper technologies as they become central to data scientists’ work. One example: Cognitive Assistance for Data Scientists (CADS) suggests, tests, and deploys machine learning models for you, so you don’t have to be an expert data scientist to build cognitive applications.

Forbes spelled out the competitive advantage the platform’s clarity and uniformity confers:

“By creating a universal platform [Data Science Experience] IBM hopes to help integrate data trapped in separate protocols residing on incompatible systems. This will not only enable more advanced analytics, it will help us to reimagine how we manage our organizations and compete in the marketplace²

IBM Data Science Experience, in function and form, help simplify the data science universe. Today, Data Science Experience is one of the premier data science systems available in the market, with thousands of users worldwide³.


David Townsend, Director of Design, IBM Analytics, San Francisco
David Schultz, Head of Studio, IBM Analytics, San Francisco
Valeria Montrucchio, Design Lead, DSX, San Francisco

DSX Designers: Caroline Law, Renee Mascarinas, Jason Azares, Jessica Gore, Leila J (IBM Analytics, San Francisco)

1. “Magic Quadrant for Data Science Platforms,” Gartner, Published: 14 February 2017 ID: G00301536

2. “IBM Announces A Universal Platform For Data Science,” Forbes, Published June 7th 2016

3. Based on download figures August 15, 2017

Arin Bhowmick (@arinbhowmick) is Vice President, Design at IBM based in San Francisco, California.