Data scientist are the people who are thought to be the statistics wizards and tech gurus.
I have written about what tech skills does a data scientist need. In most cases, these beliefs are the truth. Data scientist are expected, most of the time, to perform wonders using some fancy algorithm names and tools. Everyone is focusing on their technical knowledge and expertise, their past tech knowledge and the project they have been partly from. This is all great, it is needed and it is a big part of everyday work a Data scientist should do.
Unfortunately, not so many people focus on the soft skills data scientists should have. That is why I took the time to think about and state the top three skills data scientist should have, according to me.
Business Acumen – Have a Good Business Sense
Developing business sense is about developing a continually evolving understanding of the ever-changing dynamics of the local, national and international business climate. Good business sense often develops through hands-on experience and trial and error in the business world, although it can also be fostered through education, research and mentoring.
When you are still junior data scientist, it is the best time to use the excuse that you are new and learning new things. Like this, you can ask endless questions to different stakeholders, learn what they do, what their project is about and try to connect their projects with the whole company strategy and company’s vision.
People should not be bothered by the fact that you are proactively asking questions and being curious about how things work around you.
When you are a more senior data scientist, setting up meetings with all major stakeholders and understanding how the company teams and projects are interconnected is a natural thing to expect. So if you still haven’t done it, its the right time to start.
In the end, Data scientists are the ones who should understand the business and its processes inside and out. When critical decisions should be made based on data, Data scientists should know as much as possible about business strategy, business history, what worked in the past and what can work in the future.
It is essential for you as a data scientist to have good business sense so you can recognize specific business challenges and offer their solutions before they arise.
Persistence – don’t quit on the first bug
Not all solutions will work easily and consistently. You need to be persistent in finding a solution when the solution seems impossible to solve.
We all have been there, coding our first Hello World application. But not for everyone this application was a success from the first run.
This is where persistence comes into play. Being able to stay calm and find the error you are facing without having a nervous breakdown and too many frustrations will differentiate good from not so good data scientist.
Flexibility – be able to balance around projects people and technology
Be flexible with which environment you will use.
Not every solution can be solved in Python. Something is easier to solve in R for example.
You need to recognize how you will solve a solution fast and efficiently. No matter if it’s with your favorite programing language.
Many times you will be spread between different projects. Different projects will have different requirements that will be solved using different packages.
For example, I’m working on a few projects at the same time. I develop one project natively in R. R offers the package I need exactly to solve 100% of the project requirements so I don’t see the reason to develop the same using Python, spend enormous time and effort to develop everything manually, when I can solve it in few lines of code in R in few days.
I use native Python, to develop another project. Why Python now? I have the freedom to choose the technology because the project allows me to. I want to be good in Python as much as in R and other languages.
If I have to join all my projects together and make them accessible to the rest of the team, I use KNIME. KNIME is used as a team data science tool, so everything I do needs to end up there.
Having the skill to balance around different technologies, project and requests will make you the rockstar data scientist.