Q&A with a Data Scientist
As a graduating senior, I have had many students ask me about the BIMA major, seeking both advice and information about the data industry in general. Every major and field has its intricacies and lingo that seem opaque to outsiders. Because of the field’s many confusing intricacies, I decided that I needed the help of an expert to help demystify data science and business analytics; I sought the advice of Tom Gorin, a Principal of data science at the Boston Consulting Group — a premier consulting firm. He has years of experience at the highest level of business and data science in the world. a series of interviews about data science so that people can learn more by reading through some conversations I have had with experts in the field.
Eli Weiss: What is the value add that data science offers?
Tom Gorin: The value that data science offers is not new. First off, data science is the latest name of a set of former names for the same skillset, such as operations research groups, research and development, etc. Companies have been leveraging these skills for centuries. As was the case throughout time, the value is to find ways to do business better. That is a very wide-ranging statement. I think people usually think of data science as a way to automate things, or to focus on personalized offerings for customers (at least these seem to be the top two use cases that I hear about currently). In my mind, there are numerous other areas where data science can have an impact, from transportation to community outreach. Transportation is an easy one with Uber and Lyft, or any of the scooter companies that have sprouted everywhere. However, if you think about traditional transportation (e.g. public transportation), one wonders what they could do better if they spent more time working with their data (actually, nobody wonders: they could do much better). At the other end of the spectrum, I don’t know of any use cases in community outreach (but I am sure there are many), how would you think of addressing homelessness, for example, with data science? It’s a huge problem in San Francisco today, which is why I bring it up. I wonder (and I don’t have the answer) what we could do with data to help address this challenge.
EW: What trends do you see in the data science community that we will see over the next 10 years? What will be the impact?
TG: Increase in focus on artificial intelligence. Use cases are not fully developed, but there is certainly a lot of push towards AI. The impact is hard to tell, but it will certainly be interesting to understand how AI develops and works. I was at a conference recently where one of the speakers talked about AI in an interesting way. His perspective was that AI is not a set of algorithms, unlike what people think of today (AI = neural networks or deep learning). Instead, his point was that AI is a collection of processes and methods that start with algorithms that can be used for forecasting, sentiment analysis, etc. The AI piece comes in when you start leveraging these algorithms to automate the end to end processes that are being managed by people today. So AI is more than a set of algorithms. It’s about automation of processes that are managed by people today.
The impact of AI, in my mind, will be exactly that: the ability to automate more and more of the tasks that require human supervision and allow people to focus on value-added tasks. The idea is not to remove people from the workplace, but rather to help them focus on more relevant tasks that computers cannot handle.
EW: What are the most important skills for students to be developing right now in school to be ready for the workplace?
TG: I think there are many, but it is difficult to say with certainty. If you are thinking of data science positions, you need to build a strong foundation for data science, whether in the fields of ML, statistics, optimization or other, combined with strong understanding of the tools that you will need to apply these methods. What I mean by that is that knowing optimization is not enough anymore. You will also need to have a strong knowledge of the tools that will allow you to apply optimization, which includes (for example), Python, R, SQL, CPLEX/OPL, etc. This also means getting some understanding of the systems that companies rely on for their day to day operations. Do they use SQL relational databases, are they using non-relational databases, what are the pros and cons of each, etc.?
Photo Caption: Data science today encompasses a broad array of technical skills, including math, statistics and advanced computing.