The underpinning of data science is mathematics and statistics.  While most data science roles include some level of aptitude in math and stats, other roles require a graduate degree in one of these disciplines or PhD.  Other academic disciplines related to data science include Physics, Engineering, Economics, and Computer Science.

Python, R, SQL, Scala, and Julia are just a few of the programming languages used by data scientists.  There are also programming languages like SAS and Stata that have been around long before the machine learning models of today.  Programming languages like Java, C/C++, and PHP are used to develop software around data science-related inputs and outputs.

Whether a business has limited data science resources or not, it usually makes sense to have a data science generalist on staff.  A generalist (also reference to as a business analyst) interacts and liaises with business owners, programmers, and other data science professionals.  For this reason, such a person should have a broad range of data science knowledge as well as business acumen, but may not possess the deep skillsets of data scientists with extensive machine learning experience under their belt. 

The reality is that data scientists spend much of their time working with the data versus actual modeling.  Their work typically involves data hygiene, sourcing new data, integrating it with current data, and annotating data.  Skills to look for in these resources include automation, developing workflows, APIs, and experience with multiple data formats. Data wrangling skills may overlap with traditional IT skills such as software engineering, data architecture, and more broadly, data solutioning.