01. Research Overview
         

The Data Science Boom

Harvard Business Review has deemed the 'Data Scientist the sexiest job of the 21st century' and others have projected that over 1.5 million data science roles will be in demand as the field continues to explode.

Storytelling with Data

Data science meets a glaring need to connect people to data in order to find meaning. In that way, data scientists are storytellers. They desire to find meaning and share their insights with others.

Growing an Ecosystem

Our researchers uncovered four distinct personas revolving around data, with the data scientist as just one piece of the greater data story our ecosystem of products will eventually support. 

02. Christina, the Data Scientist
         

03. Christina's Workflow Today
         

The Big Picture

Christina works with business analysts, data engineers, and app developers. While we zoomed in on the data science portion of this ecosystem, there was an emphasis on the cross collaboration between these rolls in order to build a workspace that would support this cross persona workflow.  

A Painfully Disjointed Workflow

Right now, data scientists have to use an entire collection of disjointed tools to get their job done. Here lies the data scientist's greatest pain point, and our greatest opportunity.

04. Opportunities
         

 In this disconnect between linear tools and cyclical process there is an opportunity to bring all of what data scientists crave all into one workplace. What our data scientist, Christina really needs is to be able to:

Learn • Get started or get better with in-context learning.

Create • Leverage the best of open source tooling with IBM innovation to accomplish her work.

Collaborate • Work smarter leveraging a whole community of data scientists and work faster with her team.

05. Iteration: Collaboration & Learning
         

Building a project from the ground up in 6 months required a constant weighing of feasibility versus impact. Mapping our features back to our three opportunities helped us prioritize what to tackle. While I had a part in nearly every feature that passed this test for our initial launch, I zoomed in to focus on learning and collaboration tools as the product matured.


The following highlights the process behind several of those elements whose purpose is to enable Christina, the data scientist, to get better at what she does by leveraging the collective creativity of her fellow data scientists as she works to solver her own unique business questions.

March Storyboards

Communicating user needs at a high level to gain stakeholder buy in before diving into the details.

April Wireframes

Translating our storyboards into user flows while exploring how to fit all the pieces together.

June Beta

Balancing feasibility and impact to build a minimal viable product with a remote team of engineers.

September GA

Bolstering the community by focusing on optimizing content based components.

Community Cards

One way in which Christina can learn from her community is by exploring shared articles, projects, data sets, tutorials and more. She needs to be able to understand various content types and their relevance to her at a glance throughout her data science experience.

       
02.
"I like the thumbnail. If there's a visualization I want to know about it."

       
01.
"There's not enough here to tell me if it's useful."

       
03.
Social metrics transforming content from not only a point of learning, but a point of collaboration.

Maker Palette

The name given to our side panel, just one of the places the above community cards live, so that Christina can be coding in her notebook to the left and referencing material on the right. It is a palette of knowledge, enabling her creating, and allowing her to 'paint' with the work of others. This interaction underscores her desire to experiment and iterate in tight loops by bringing all of her materials into one workspace.

       
01.
Initially, the maker palette included a proposal for a scrubbing panel to optimize the side by side workflow and account for many content types. Due to feasibility, the interaction was de-scoped and saved for later.

       
02. 
Feedback continually highlighted the struggle of limiting real estate without overwhelming the user without content overload.

       
03.
The final iteration of the maker palette included visually breaking content types up into bookmarks, help, and community to help organize a library of possibilities.

Bookmarking

Enabling Christina to not only explore and find relevant content, but save it for later so that she can fully leverage the information she uncovers. The biggest challenge for bookmarking revolved around scaling the interaction to match different card types and sizes without having to rebuild a library of components.

06. Outcome
         

 Since the launch of Data Science Experience in 2016, the product has been rebranded as IBM Watson Studio, but much of the foundation I was a part of creating still exists. On top of winning the 2017 Red Dot Award for Communication, the original Data Science Experience made a big splash in the news with outlets such as Forbes stating "This will not only enable more advanced analytics, it will help us to reimagine how we manage our organizations and compete in the marketplace."

01.

Now data scientist Christina can
get educated

Christina can get better at what she loves to do by watching tutorials, reading documentation, and troubleshooting while she codes.

02.

Christina can
document her progress

Through the use of projects, Christina can package all the assets she used through the course of her assignment to refer to later, pass off to colleagues, or share with the community.

03.

Christina can
collaborate with peers

Christina can work across organizational silos by sharing projects with her team for quicker results and business impact.

04.

And she can
join a community

Christina can enrich and be enriched by a community of fellow data scientists to further the field.