My Journey Into Data Science

Photo credit:

I graduated from college in 2008 right in the midst of the financial crisis with a major in psychology and math, subjects I found fascinating but hadn’t chosen based on potential career paths. Feeling at a loss of what to do, I embraced my youthful idealism and decided to volunteer in Gonaïves, Haiti in the aftermath of Hurricane Hanna. In Gonaïves, I worked for an organization called Hands on Disaster Relief and helped dig houses out of the mud and rebuild wells. The work was rewarding (and at times back breaking) but didn’t lead me any closer to a potential career. After returning to the states, I continued searching for a job that would both use my education and give me a sense of meaning and purpose.

I started a Masters in Public Health as a way to continue the path that I had started in Haiti, but realized that I was more interested in applying this to work rather than learning the theory. I ended up at a small startup focused on increasing how well and systematically patients and providers could be matched. I had very little technical experience at this point (I’d taken one C++ class in college that was at that point a distant memory!). But I was fascinated both with how we could solve this one healthcare problem, as well as all of the other problems surrounding the healthcare industry.

While working at this company, I got limited experience working with selenium scripts to run some automated tests. While most of my work involved modifying existing scripts (and making good use of copy/paste functionality!) I really enjoyed the satisfaction of writing something that would execute to exactly what I wanted to do. I still didn’t have enough technical experience to understand how I might one day use this, but it definitely planted the seed of interest. In parallel, I was gaining a deeper understanding of the healthcare industry in general, and it’s multitude of problems.

One particular area of interest was interoperability of different electronic medical record systems. Naively, this seemed like an easy problem to solve. All you needed to do was have one unique identifier for a patient and you could pull all of their disparate records across all different systems. However, when you start diving into this problem, you begin to realize that “one identifier” is not as easy as it seems. You can assign a patient a medical record number, but of course this is unique only to the system in which it was assigned. You can use other combinations of characteristics but these all begin to break down at certain points.

I continued in the healthcare tech startup world as a solutions architect/technical project manager, responsible for helping healthcare systems implement software. This was interesting to a point, but I was beginning to realize that I wanted to be more focused on actually diving into the data itself as opposed to the problem solving around it. Intrigued by the initial question of health record interoperability, I wanted to explore in more depth this world of health care data. I began to wonder what other questions I could be asking once I had more tools with which to explore.

This brought me to the General Assembly data science intensive program. This seemed like the perfect crash course to fully immerse myself in data science techniques. I am apprehensive about the speed of the course (and the fact that I am working on a Pokemon lab on a beautiful Saturday in June) but I am excited for what is to come and the myriad doors it will open.