If data is the gold of the 21st century, then data scientists are the miners — panning, extracting and processing, then transforming it into value.
The field of data science is still relatively new; companies increasingly continue to incorporate digital data collection into their business strategies. The position is often traced back to 2005 when the National Science Foundation recommended developing a career path for data scientists to ensure that there were enough experts to handle the sheer amount of data being collected.
This massive amount of data is revolutionizing nearly every industry, impacting nearly every aspect of our lives — commerce, healthcare, transportation, you name it. The career outlook for data science is bright, with the U.S. Bureau of Labor Statistics projecting that employment in the field will grow 21 percent from 2021 to 2031, much faster than the average for all occupations.
However, since the field is still finding its footing, it’s hard to know exactly what to expect when pursuing a career. That’s doubly true for those just embarking on the job hunt for data science roles.
Since the data science field is still finding its footing, it’s hard to know exactly what to expect when pursuing a career.
“Data science roles are relatively new. Because of this, the field hasn’t settled on a standard set of job requirements or responsibilities yet,” said Quant Research and Development Manager Bian Elinsky, of Belvedere Trading in Chicago. “So a data scientist at one firm could do completely different work than a data scientist at another company. To further complicate matters, companies have many names for jobs similar to data scientists. At Belvedere Trading, the quant trading analyst and quant roles are most similar to the traditional data scientist role.”
For more insights into what to expect when applying for your first role in data science, Built In Chicago talked to Elinsky for tips and tricks that early career data scientists can use to land a job.
Belvedere Trading is a proprietary trading firm.
Tell us about the candidate interview experience for a data science role.
Most first-round data scientist interviews will cover both behavioral and technical interview questions. The technical questions can vary from statistics to specific modeling techniques, to Python language questions. Because this range is so wide, you will probably miss a few questions, or have suboptimal answers. I’ve learned not to put too much weight into the result of a single interview.
Take notes on the questions they asked, and what you missed. Practice answers to what you missed. Over time, you will get better at interviews. And you will have a better idea of what types of questions get asked in interviews.
What is the most important thing you’d recommend to prepare for a data science interview?
The wide variety of job requirements and responsibilities in data science makes it especially difficult to prepare for interviews. Required skills include probability, statistics, machine learning, modeling, programming and algorithms, SQL and data manipulation, math, and product questions. You can’t become an expert in every area.That makes it especially important to identify your strengths and then apply to companies that value the skills you are best at.
The wide variety of job responsibilities in data science makes it important to identify your strengths and apply to companies that value the skills you are best at.”
Don’t apply for all data scientist roles under the sun. Find your niche. Practice your strengths. Make sure your weaknesses don’t become debilitating. Target companies and roles that are great fits for you.
What advice do you have for someone preparing for a data science interview at your company?
At Belvedere Trading we work with financial datasets. Most financial data is time series in nature. It is especially important to brush up on time series data analysis methods and forecasting models.
Be sure to have a clear answer to why you want to work in trading. Candidates that are intrinsically interested in options and financial research are stronger than those who aren’t. Knowing basic option pricing theory can also go a long way. We are an options market-making firm, not an equities market-making firm. So don’t tell us what stocks you think are long or short!
Options market making is also very mathematical. It is important to be good and quick at basic probability and statistics concepts. It’s common for quant trading analysts at Belvedere to do quick back-of-the-envelope calculations to make quick estimations.
At Belvedere, we value relatively simple explainable models over complex, less-explainable models. If you want to show off skills in more complicated ML methods like dimensionality reduction or deep learning, be prepared to explain your ideas clearly and concisely, and come prepared with explainability methods, such as feature importance for decision trees.
