Past Event

Data-Science-as-a-Function: A Conceptual Framework

March 31, 2022
11:00 AM - 12:30 PM
Event time is displayed in your time zone.

Virtual lecture by Marco Morales

Visiting Assistant Professor, QMSS, and Former Data Science Executive at Warner Music Group and NBCUniversal

Thursday, March 31st, 11am (virtual - RSVP required)


The Quantitative Methods in the Social Sciences (QMSS) MA program at Columbia is delighted to host a lecture by Visiting Assistant Professor, Marco Morales. 

Dr. Morales will be speaking about the evolution and current state of Data Science in business organizations. 

It will take place virtually on Thursday, March 31st at 11am.  This event is open to all, but it does require an RSVP in advance here.

Dr. Morales will discuss how Data Science is traditionally understood as an activity performed by individuals with specific skillsets. Because of this understanding, we have spent a lot of time and energy addressing  what Data Science should be and what defines a data scientist (e.g. skills, tools, activities).  

Given how Data Science has evolved in practice, we can – and undoubtedly should – rethink Data Science as a function that exists in industry. This is an evolving function that takes different shapes depending on the particular characteristics of the firms and industries where it exists.

At minimum, a complete conceptualization of Data-Science-as-a-function – which Dr. Morales calls the Data Science Shop – clarifies:

  • What does the function do? (i.e. the fully-fledged data product cycle)
  • Who crews the function? (i.e. archetype personas)
  • How does the function operate? (i.e. interaction between tasks of archetype personas)

Many benefits arise from evolving our thinking to this data-science-as-a-function framework:

  • Education: What we need to teach (and potential gaps in current offerings)?
  • Hiring managers: Who needs to be hired for the function and with what skills?
  • Executives: Is a firm mature enough for the function?
  • Diagnosis: Why does the function succeed and fail at different firms?

This talk is intended to lay the foundations for this evolution of our conceptual thinking on Data-Science-as-a-function.

Please join us for this lecture.  RSVP in advance here.  

Dr. Marco Morales has taught Applied Data Science for Social Scientists and Theory and Methodology at QMSS since 2016. He is currently in residence as Visiting Faculty during the Academic Year 2021-2022. During this time, he is continuing his work on identifying the conditions that ensure success and adoption of Data Science in corporate environments. Dr. Morales brings an incredible wealth of behavioral science knowledge and executive experience leading Data Science organizations in the media and entertainment industry, most recently at Warner Music Group and NBCUniversal. Throughout his career in Data Science, he has led teams of data scientists and data engineers to build automated platforms and data products tailored to the media space. He received his Ph.D. in Political Science from New York University His current research interests include applications of Machine Learning for Inference, Experimentation and Causal Inference, and the formalization of the differences between inference and prediction.