There are probably dozens of variants of the Venn diagram that Drew Conway proposed a few years ago to capture the core skills of a data scientist. Needless to say, the role has experienced many changes since then, while rapid technological developments and the boom of AI have further propelled the profession to the top of LinkedIn’s emerging jobs ranking.
Well — we couldn’t resist putting forward our own version of the infamous Venn diagram. Like Conway’s, ours is built on three axes. However, our model focuses on broader categories rather than on specific expertise. In today’s ever-changing business world, soft and cross-cutting skills are the truly decisive factors that, in the long run, can ensure adaptability and success.
Thus, our “holy trinity,” if you will, of data science is made up of:
- Curiosity
- Technical know-how
- Collaboration
Thinking of a career in the field, or wondering if you’re doing this right? Let’s dive into each component.
The importance of a curious mind
Probably obvious, but it’s impossible to talk about science and not mention the innate curiosity that powers it. Whether you plan to explore the possibility of life in other planets or the mysteries of quantum entanglement, it is the thirst for answers to questions and riddles that will make you advance.
This, of course, applies to the problem-solving capabilities required in data science projects. Nevertheless, well-directed technical inquiries tend to fall on shaky ground whenever there are not accompanied by a good contextual understanding. Just because you’re good at playing with data and creating models that produce intricate insights and machine learning experiences, none of it is worth anything if your work isn’t helpful to the overarching goal.
For this reason, the need for curiosity expands to the domain of expertise in which you operate (i.e. finance, political studies, marketing). The more you know about the field of work of your company or department, the better questions you will ask yourself, the useful insights and models you will produce.
Note that we’re highlighting “curiosity” rather than “knowledge.” You’re going to spend many hours working with this data. Make sure it’s something that you are passionate about or at least find interesting.
Knowing the technical ins and outs
Some describe a data scientist as someone who knows more about math and statistics than your average programmer while having greater coding capabilities than your average mathematician. Although this definition errs on side of oversimplification, it is not totally misguided.
To be successful in data science, you need to be proficient in certain data engineering and coding-related methodologies and practices. It is important not only to know how to build effective code, but also how to efficiently extract and clean data.
Additionally, there is the crucial technical knowledge that has less to do with computer engineering and more with, for instance, data privacy compliance. You must know what data sets you can manipulate and which ones you can’t, which processes can be computed on the cloud and which ones are better reserved for on-premises infrastructure. At the same time, if you work in finance or in any other field where sector-specific concepts are a basic requirement, you will have to dominate those on top of your knowledge of data science.
Playing as a team
This is where soft skills play the biggest role. Interpersonal communication and teamwork have always been one of the key factors of success Their relevance in this hyperconnected world of ours is only increasing.
There must be good cooperation between all teams and stakeholders involved in the process, and, for that, you should be able to communicate efficiently and in a compelling way. It’s not enough with working closely with developers or analysts. Knowing how to present a project in layman’s terms becomes essential if you want to be granted the staff or computational power that you’ll need to complete it.
Apart from this, you need to be well-versed in concepts like Agile development, which help teams streamline the production pipeline. Version control, a unified repository, and a good understanding between development and production are a teamwork-must in today’s IT world.
