Data Science May Never Be the Same

September 7, 2016
· 2 min read

Snap Advances’ Chief Data Scientist Justin Dickerson recently published an interesting LinkedIn post on the future of data science — explaining how modern tools like DataRobot are helping data science teams scale their operations to meet business demands, while preserving the elements of the function that data scientists enjoy most.

“If you take the model selection and optimization process out of the data science workflow (because you’ve automated most of it), then you’re allowed to focus on better defining the business problem, ensuring you have the right data to answer the right question, and becoming a much better feature engineer (which many of us would argue is the most important part of machine learning anyway).”

Top three takeaways from Justin’s article:

  • Time-consuming parts of the data scientist’s job — tweaking inputs, feature engineering, optimizing parameters, implementing the models, maintaining quality assurance — aren’t what makes it enjoyable for most, nor do those tasks facilitate data science at scale, which is a key priority for the business. It can prove difficult to develop an industry-leading data science team that is challenged to innovate and truly enjoy the art of the craft while meeting business demands for revenue and operational efficiencies at scale.
  • The DataRobot platform facilitates and automates the data science tasks that prevent scale, incorporating best practices from top data scientists and building in workflows to optimize machine learning success.
  • It’s hard to predict what data science teams in the future will look like, but they’ll probably have closer working relationships with business teams and more bandwidth to solve business problems.

To read Justin’s full post, visit the article on LinkedIn.


New Call-to-action

About the author
Chuck Smith
Meet Chuck Smith
  • Listen to the blog
  • Share this post
    Subscribe to DataRobot Blog
    Newsletter Subscription
    Subscribe to our Blog