Team Meytier is excited to share this guest written blog by Afrozy Ara, director of Data Science & Analytics at Incedo Inc. Afrozy shares her unique perspective as a hiring manager on what makes a good data scientist and the important skills and qualities her and other managers seek out.
Despite the disruption of the Covid-19 pandemic, the demand for Data scientists continues to explode. In 2019, LinkedIn ranked “Data Scientist” the No. 1 most promising job in the U.S. based on job openings, salary, and career advancement opportunities. This trend has strengthened in 2020 as Data science is booming and starting to replace legacy roles.
As a Hiring Manager in this Industry, I see some very paradoxical sides of supply and demand. On one side, I come across many who want to build a career in Data Science being enticed by the opportunity it offers and the “sexiest job of the century” label. On the other hand, I find it very difficult to find good talent in the industry - the interview funnel has a very steep drop-off. Given the huge demand, there is a stark need for getting more Data professionals into the talent pool, and one of the best ways to make that happen is reskilling those who have an interest in this field.
The Data Scientist though, is a hybrid creature - someone with expertise at the intersection of math, business and technology. Hence an unintended consequence of it is that this has become an abstract discipline which is understood well only by those who have been in the weeds of it. It also tends to get confused with adjacent disciplines like Data Engineering which is more about building the pipelines and plumbing that makes Data Science work. To add to it - the hype around AI does not help.
But this is not an unsolvable problem. If you are looking for a way to pivot your career as a Data Scientist, you can build your profile and develop the skills needed to get the job. In this article, let's unpack the thought process of a hiring manager and the skills they look for while hiring for Data Science.
Problem solving capability: This means that you have the ability to take any business problem, break it down into its constituent pieces, frame it right and solve it. Data scientists need to be able to look at the desired future state and define what is the problem that really needs to be solved. Most times the answer that your stakeholders think they want is not the answer that they need, hence having the ability to think clearly is unmistakably the most important skill, especially for senior roles. The thing about problem solving is that it is something that you learn gradually, through practice and observation. Consciously develop it in whatever field you are working on because shining here is a sure shot way to impress your interviewer.
Basic technical skills: A vast majority of the work for data science would be understanding, exploring and extracting insights from the data. The absolute basics are SQL, Python , Statistics & Machine Learning. And being comfortable with visualization tools like Tableau also helps. Over here, don't get carried away by the hype and start with the data munging and exploration step first. Deep learning and Neural networks - though have the fancy zing to them, would come in much later. Getting the technical skills right is the easiest capability to build so if you are thinking of switching into a career in data science, this is a great place to start.
Communication and storytelling: Another key aspect of the Data scientist is being able to influence decisions without authority and through collaboration. As someone closest to the data, you have exposure to the truth of what the data is telling us. However, the truth can be complex - so how do you translate it in a form which non-technical teams can understand? Other times, the answer you get is not the answer that the business was expecting to hear. Communicating effectively to executives so that they understand patterns or truth the data is revealing, irrespective of what the original hypothesis was is also a key skill Data Scientists need to develop to be successful in this role.
Understanding Insights: this means being able to look at the data and extract insights from it, regardless of what platform or technology you are using. And connecting that Insight to the problem to eventually make a business impact. Lack of understanding of Insights is a key gap I observe in many who are attempting to crossover to the Data Science side. Having a strong domain and data orientation is an important way to overcome this gap. Take a dataset and pry it apart - what is the story that this data is telling? What are the insights you can generate from this data? Are there any inherent biases in this dataset? Every domain has a unique shape and characteristics of the data and what insights mean in that particular industry. Being able to understand the nuances of data, and how it would translate to business impact is critical for any Data science role
Data Tooling: as companies increasingly move towards analytics platforms and away from ad-hoc solving, this is something that would differentiate the good and the exceptional. Understanding the data tooling for full stack analytics and personalization capabilities and being someone who can bring together the Data Insights and the Engineering aspects of it would be super valuable. There are typically two approaches to Data tooling i.e All in one platforms like Amazon Sagemaker or best-of-breed products like Snowflake, Kafka etc which are focused on one aspect of the Machine Learning process. Having an understanding of the end-to-end pipeline will help you wear many hats and is a great advantage for Data Engineers moving into Data Science.
This advice is targeted towards those who want to get into the technical role of a Data Scientist. If you are not technically oriented yet want to develop a career in Data, don't lose heart. There are many non-technical roles in Data which are equally critical and impactful.
I wish you the best of luck in your Data Science journey!