Career Paths Within Data Science
Data science is a broad field. It may be that due to the massive amounts of data that can be collected and used in nearly every industry and in countless ways. It could also be because it combines many different areas of expertise — math, stats, programming, domain expertise. It may even be because the topic is new enough that all we have are extremely broad and varying definitions to describe it in all its forms.
Big data is pervasive, in industry and beyond. A grocery chain keeps track of all the different types of food they sell, as well as how much and the price per good. A bank keeps track of its customers, each account they open, how much money is in each account, how much debt they own and much more. Is all of this information useful? Well, I’d like to think that some of it is so this whole thing wasn’t a waste of time.
Even the more obscure jobs which people don’t normally associate with data science could use big data: hippotherapists use horse riding for physical, occupational, and speech therapy. Here, data can be collected on each patient and each horse to create models for what horses, what exercises, what speeds, what heights, ect. will help patients with certain problems based on data from past treatments.
Due to the various ways different industries handle data and put it to use, there have been many different job titles created for those who use data on different levels. Here, we’ll go through each of the primary job titles and their descriptions in salary order according to Glassdoor (we all got into this, at least a little, for the money).
Job title: Data Analyst / Business Analyst
Skills and goal: Use code (often but not always python) to create data visualizations. Presenting data in the form of histograms, scatter plots, heat maps, etc. helps us to understand trends and correlations that might be useful in making tasks more efficient and effective. May occasionally perform some hypothesis testing or other statistical methods to justify correlation.
Example: All aboard the Stewart Hall elevator in Warren College of UCSD, perhaps the slowest moving entity to ever enter the campus. Not only can it go down, but it can also go up (more than I can say for my GPA). Imagine a data set containing all the times this elevator was used: the floor someone got on, the floor someone got off, the time this took place. Wow. Incredible. But not all things UCSD must be bland and boring (as Khosla might have you believe). We have the power to make these numbers into pretty pictures known better to data analysts as data visualizations. Perhaps a bar graph to depict the frequency each floor is visited, or maybe a scatter plot for ascent and descent times throughout the day. We have ways to depict the data which make it easier to spot trends, like the elevator taking a long time always, that may help us make improvements (unlikely), or better decisions for efficiency (stairs).
Average wage: $62,453
Job title: Data Engineer / Business Intelligence Architect
Skills and goal: Fit for computer scientists, data engineering is heavily programming-focused as they process data from when it is collected to it’s transformation into a data set that can be cleaned and analyzed, also known as creating a data pipeline.
Example: Remember that data set we imagined before so we could make visualizations? Well thanks to data engineers, our data sets don’t have to be imaginary. Say, for instance, that in an elevator as fast as a car without an engine, someone with solid software engineering skills decided to incorporate, into the computer that controls the elevator, some software they developed that would keep track of such information as each floor the elevator visited, the times it did so, and the weight it carried. They have the computer track this information and then format it in a way such that data analysts can work with it. Sounds simple but the development of the software can be tough.
Average wage: $102,864
Job title: Data Scientist / Machine Learning Scientist (Broad Title)
Skills and goal: Make models, usually predictive and/or implementing machine learning. Requires programming as well as higher math (such as calculus and linear algebra) and statistical analysis skills. Machine learning scientists are data scientists who create new algorithms using the same kind of skills with a focus on machine learning.
Example: Previously, we learned that our elevator data set could be used to make visualizations that could reveal trends that would not have been apparent otherwise. Now, we go a step past data analysis. We try to use the data from past runs of the elevator to predict what will happen next. Perhaps we notice that 90% of the time that the elevator stops on the 5th floor, it’s next service will start on the first floor. Maybe we almost never see the elevator take trips that start on the 1st floor and end on the 2nd. It could happen to be that during a particular quarter, the elevator is always called to the 3rd floor at the same time every Monday, Wednesday, and Friday. All this is information that could be used to make future runs more efficient. It won’t, but it could.
Average wage DS: $113,309
Average wage MLS: $114,121
Job title: Machine Learning Engineer
Skills and goal: A combination role of data engineer and data scientist whose goal is to take a model (like what a data scientist would create) and deploy that model in industry, which may require rewriting the algorithm, transforming the data results into usable procedure (which requires the skills of a data engineer).
Example: At last, the hard work can be put to use. With some more hard work, that is. A machine learning engineer could be employed to take the predictive model for our favorite elevator, created by the data scientist, and rewrite the algorithm into the elevator’s program so that, after finishing a job, it goes to the floor it has the highest chance of being called to next according to the predictive model. This job requires the knowledge of both the data scientist and the data engineer because they have to understand the data scientist’s model, be able to rewrite it to be usable, and install that software onto the computer of the elevator. Truly impressive, especially considering the technologically obsolete elevator isn’t capable of anything described in this article.
Average wage: $114,121
That concludes the primary technical data science career titles according to my opinion. Before closing out, however, I would like to mention a few less technical/more creative or business-oriented positions that also have a focus on data science, keeping in mind the versatility of the field.
User Experience (UX) Researchers seek to gather data on user experiences so that companies can improve their product design based on their findings. Naturally, when we start so see larger quantities in gathering and interpreting the user feedback, data science is used to make the information easier to understand.
Sales Engineers work to sell technology that will increase sales for another company. Because the product they sell is typically software that uses data science to increase efficiency of sales, they have to have a basic understanding of the algorithms and analytics behind these processes, as well as how the typical sales of the company function.
Conversational UI Script Writers can come from various backgrounds such as script writers, bloggers and behavioral psychologists. Their goal is to make bots sound more human-like and even develop an artificial personality.
There are many more positions which I could write out but that list would be unnecessarily long and redundant. Thanks for reading and good luck with the path you decide to take!
