Is it our future to be agnostic?

As interest in data and data analysis grows, future students interested in the career have to work harder to understand the boundaries and guidelines. It won’t always be as simple as it was for Evan Thorne ‘15, who came to Wesleyan thinking he wanted to study Economics before discovering the QAC department. “I was starting to work with data sets in my math and computer science classes when I heard about big data,” Evan said, explaining that it was soon after he began taking classes with the QAC that he realized he wanted this as a career.

But what is “this”? Data science? Data analysis? Data manipulation? Sometimes it can be hard to define. But Evan did not flounder when I asked him for a definition of his job at CKM Advisors, the company where he was hired right after graduation. He began by explaining to me that an analyst is someone who is able to take in what’s readily available to them and then dissect it to look at more basic stats and trends. Data scientists, however, are able to find things that aren’t available – unstructured data – and take it in raw. “Every data scientist is an analyst in a way,” Evan explained, “but it’s at a much bigger level.” At CKM, Evan is a data scientist, and he is responsible for all of the analytic process: data ingestion, wrangling, manipulation, analysis, and visualization.

It seems to be a growing expectation that incoming data scientists need to be able to cover the entire pipeline. Those doing this work must be flexible in their sources and adjustable in their methods, putting the value of mastering specific languages below learning how and why everything works. This poses an interesting question to the future data scientist: Is the best option to ditch favorite technologies in favor of being completely agnostic?

I asked Evan about the projects that he’s worked on. While he couldn’t say much due to non-disclosure agreements with the companies, he said he’s worked in real estate optimization, financial services, and company operations. In order to work on these issues that are specific to the companies, Evan is given access to their data. Unfortunately, these are not data that any student could access. He has to go through intense security in order to log in remotely from his New York office.

“We then dive into them using Python,” Evan explained, which led me to ask him about CKM’S declaration on their website that they are “technology agnostic.” Evan explained that this means when a data scientist is tackling a problem, they can tackle it in whatever language they feel most comfortable with. “It’s about the result,” he stressed. And while CKM strongly recommends Python and R, Evan wasn’t as familiar with Python when he came in – but now he uses it every day. This supports the agnostic hypothesis, that what these consulting firms are interested in are the skills and knowledge, rather than one specific language.

This might seem counter-intuitive in an education system of standardized testing and knowledge maximization, but clearly it’s very important to learn this flexibility. “The industry is changing really fast, so it’s about solidifying the foundation of data sciences,” Evan said. I pointed out that this can start at the QAC. Evan agreed: “The new libraries are really changing a lot. So [The QAC] would need to teach the basic data frames and fundamental functions for any type of analysis, no matter what tool is thought of the best.” As for generally insuring that the QAC can teach us the best analysis skills, Evan wasn’t without ideas. “I think predictive modeling would be a good class. That’s important for a lot of the data science work that’s currently happening. Everyone wants a forecast.”

So, what’s our forecast? As prospective job hunters who would hope to be like Evan, how do we prepare for an industry that is growing and changing how we speak? It seems like there is a way to have our bases covered, to not be off guard. We can focus on how things happen instead of how we’re specifically making them happen. If we have the skills, if we stay agnostic, then hopefully we could be customized for any opening that comes our way.