This is the first in a series called #VisualizingData, which highlights unique professional profiles in data science. These stories are such an important part of the narrative around stereotypes and opportunity in our communities, and I hope you’re as inspired and encouraged as I was - to see parts of yourself in these stories and become aware of the wide range of paths to a career destination.
Alexandra Siega is a Language Analyst turned Data Scientist in New York! I met her through the community of students growing their skills with the Udacity Data Analyst Nanodegree. I was so excited to speak to her because her background and journey to data resonated with me, and she lived up to all my hopes by being extremely authentic in her answers. Hope you enjoy! (all emphasis mine)
What does data science look like at your current organization? How do upper level execs use your predictions?
I work for CKM Advisors, a small, data-driven consultancy in NYC. More than half of CKM employees are data scientists, with the rest being “data-adjacent,” meaning that they’ve picked up coding skills and directly contribute to the analysis process. It’s a very flat organization; thus, the concept of “upper level execs” only exists externally through the clients we serve. As we’re directly communicating with clients on a regular basis – even us data scientists – we have great insight into how our analysis is used: largely to identify and rectify areas of inefficiency.
At CKM, a data science project spans the whole analysis process: from first finding and ingesting the data, to presenting final insights to the client. There are a number of competencies that we’re expected to develop over the course of our careers as data consultants: dashboarding/web development, machine learning, data munging and analysis, and business relationships, to name a few. With multiple competencies as the focus, this means that a typical day as a data scientist at CKM often varies. One day, I could be knee-deep in a classification problem; the next, I could be working on adding new D3 visualizations to a Django dashboard; still another, I could be on the phone with the client discussing findings. I personally love the challenge of drawing upon all of my skills during a project – especially as someone who is coming to the field from a non-traditional background.
## You completed a 3-month, full time web development bootcamp with the Flatiron School before moving into your data science role. Can you share about that transition? Did you go into the boot camp knowing you wanted to eventually move towards data science? Was it difficult to convince your current employer of the transferability of your skills?
I specifically participated in a web development bootcamp to get away from data, actually! As a language analyst for the government, my main interaction with technology was through processing data: processes that I desperately wanted to improve by developing more effective web tools than the ones available. Once I started down the path of web development, however, I realized that the web debugging process was not fulfilling. Yes, there were problems to solve – chaining AJAX calls or implementing authentication, for example – but they were known problems, and the solution was always lurking in a Stack Overflow post somewhere. On the contrary, data analysis is all about making sense of the unknown; Stack Overflow can’t answer the “why” that is so essential to problem solving in the field.
With my former background in analysis and demonstrated interest (this site, SectorSalaries, was my final immersive project at the Flatiron School) in doing data work even while still immersed in web, I found that crafting my pitch to do data science professionally was relatively straightforward. There’s a need for people who span both web development and data science (e.g. D3 visualizations, dashboards), so convincing data-focused organizations of your usefulness isn’t particularly difficult!
How do you use skills from your language background, in your current data science work?
I’ve found that my approach to language analysis and translation is pretty similar to how I solve data problems. When confronted with a new dataset or foreign language text, as both a data practitioner and translator I strive to understand context before I even start any nitty-gritty analysis:
For me, the analytical process is abstracted away from the subject matter. Whether I’m working with language or data, I’m always going from the broadest to the most detailed level of insight over the course of the analysis. Only once I have context am I comfortable tackling the question or problem at hand!
## What was the interview process like for your current role?
Beyond the typical culture/fit interviews, CKM has two additions to the process for data scientists: a data challenge and a case study. The former is a take-home exercise that is meant to assess your technical and analytical skills. We’re tech agnostic at CKM, so we’re not looking for the Best Python Code Ever; we’re looking to see how you implement the tools you know to produce clean, accurate analysis. In a similar fashion, the case study helps us understand how you reason through an unfamiliar problem – however, it’s delivered in person.
## More personal now: Did you have any hesitations about your ability to transition during your journey towards data science?
Often non-STEM work requires lots of strategy and ‘big picture’ type thinking, and it can be hard to image working with the details full-time. Did you experience this? If so, how did you overcome it?
Hah! Self-doubt and imposter syndrome are absolutely real and rampant, and something I’ve yet to overcome. Relatedly, I don’t really view my transition to data science as complete – it’s an ongoing process in a growing field.
What has helped ease the difficulty of the transition is decoupling technology from analysis. It’s one thing to not know a language or package/library; that’s an easy to overcome with focused learning. (Fun fact: I arrived at my data science job not knowing a lick of Python, and now I use it daily!) What’s harder is developing a strategy for thorough, accurate analysis: understanding when to use the tools you’ve learned for maximum impact. Give it time; I’m certainly giving myself a few years!
## Any words of encouragement for other non-STEM professionals moving towards data science?
There’s room for so many types of people in data science. It’s a field that rewards curiosity above all – true to the “science” of it, you get to test the assumptions you develop at every level of the problem. In my opinion, the most important trait of success in the field is not necessarily being good at bleeding-edge machine learning technology, or being a hotshot at parallel computing and big data; it’s being able to contextualize an issue and anticipate and overcome barriers to a solution. Every day, I strive to stay curious and creative; these are traits that we all have within us.