Perspectives from Two Data Scientists

Aaron Burciaga from HCL and Sivan Aldor from Weelio discuss data science and analytics with Ann Beheler from CTC at WASTC 2020.

How is your program preparing students for the coming boom in data science and analytics?

The National Convergence Technology Center (CTC) recently hosted a breakout session on data science at the WASTC Winter ICT Educators’ conference.  We welcomed Aaron Burciaga from HCL and Sivan Aldor from Weelio to share their perspectives on the data science industry and how programs can prepare students.  Below are some highlights from their discussion.

* The time is ripe for entry into the data science field because it’s not yet a “well-defined career.” The workforce comes from a variety of backgrounds and academic institutions aren’t yet sure how to teach it. This means now is the best opportunity to get into the field.  In ten years, once schools catch up, entry-level positions won’t be open without formal training.  Get your foot in the door now.

* In looking at new hires, data science organizations care less about schools and programs and more about demonstrating experience.  Has the applicant done a project with real data, not studied theory in a classroom?  Has the application participated in a data science competition like Kaggle?  Has the applicant worked in the business world?  Even jobs in sales or customer service can demonstrate an ability to communicate and interact with others.  It’s not enough to have a “deep understanding” of data science; applicants need to have practical experience problem-solving with data sets.

* To this end, the “biggest gift” educators can give their data science students are actual, real-world industry problems to solve.  This should be a three-month project.  For example, how many agents will be required in a call center to provide a specific level of service?  How long will it take a pharaceutical company to develop a new drug?  How can a police department best distribute officers to appropriately reduce crime to a specific level?   Maybe the school can approach local businesses to find out a data analysis might be able to help their business.  Most often, this means helping the business develop a better customer profile. Rather than pay $1 million to a data science firm, students can provide the analysis for free using open-source software.

* Students need experience working with messy, incomplete (and sometimes irrelevant) data. They have to make do with the data they have.

* Students need to also learn the value of talking to the those who want the solution. Often, they’re not even sure of the problem they’re trying to solve.  The data scientist must ask questions to better understand the needs of the customer.

* Data scientists don’t often use advanced calculus, so it’s possible educators need to rethink the math requirements for data science programs.  Most often, calculus is only needed when creating a new model.  Educators should consider doing as much as possible to add meaning and context to math, rather than focusing just on the rules and the numbers.  Make the math mean something.  More important than a year of math schooling are skills like logical thinking, coding, and communication.

* Java, R, and Python are the standard coding languages in data science.  The standard data analytics software programs (for now) are TensorFlow and PyTorch, but there are many diverse programs and softwares. More senior level positions would expect experience in the visualization tools like Looker and Tableau.

* Aaron advocates a new way of looking at data science entry-level positions called “Blue Collar AI.”  This includes data “farmers” that grow data, data “miners” that extract useful data, and data “welders” that create apps to connect data to the business needs. Skills for these kinds of positions could be taught in a two-year degree program.  High-level executives cannot do their job without this kind of support.

* Educators should also be aware of the industry’s professional certification organization.

If you’re not aware of this WASTC Winter ICT Educators’ conference, mark your calendar now for the January 7-8, 2021 edition.  This conference runs in the Bay Area, typically costs only $100 to register, and covers a wide range of cutting-edge IT topics and programs for educators. Learn more about the 2020 conference here.