Data science has been a hot phrase on the job market in the last several years. But what does data science really mean? What are the job types waiting for you?
What Is Data Science?
The reason it’s difficult to define it easily is that it’s actually the amalgamation of several disciplines. When we talk about data science, we think about programming, mathematics, and statistics merged into one discipline.
In data science, all these disciplines (along with some others, depending on the job specifics) are combined together to study data and solve real-world (business) problems by getting insights from data. It’s a way of looking at historical and current data to predict the future.
What Are the Qualifications and Skills?
While there are no strict rules on ending up in data science, some educational fields move you closer to data science more naturally. All the skills required couldn’t be covered by such an article. However, some technical skills and knowledge are almost mandatory if you want to work in a data science field. We’ve tried to break this down in the table below:
The Jobs Overview
In the table below, there’s a job overview so that you can wrap your head around all the job opportunities. Along with the job title, there’s an overview of the most common programming languages, tools used, and the field of data science the job is focused on. Last but not least, there’s an average salary column. Of course, you’re interested in how much you can earn.
Now that you got the jobs overview, we’ll go through all the job descriptions in a little more detail.
As a data scientist, you’ll manipulate and analyze data. Then you will use it to solve business problems by building machine learning models. The purpose of all this is to provide your stakeholders with insights into data and its trends. Predict the future in a way, so they can make business decisions with more confidence in the desired outcome.
The main job of the data analyst is to conduct, sounds logical, the data analysis. To do that, they have to be versed in data querying, as well as mathematical and statistical principles to provide a good data analysis and metrics calculation. Not only that, but they also have to present their finding in the form of eye-pleasing data visualizations and dashboards.
The data engineers ensure that data scientists, data analysts, and all other data users get their data on time and in the format that is suitable for their job. They do that by focusing on data infrastructure development and maintenance, including extracting, transforming, and loading data.
If you work as a BI developer, you’ll be focusing on designing, building, and maintaining graphical dashboards in the BI tools. You’ll also perform data analyses and ensure all the automated reports and dashboards get fed with data and be regularly updated.
The business analyst focuses on the business processes and systems within the company. They analyze them and try to improve them by changing the processes, which usually means introducing new systems or improving existing ones.
A marketing scientist is a data scientist focused on marketing data and solving marketing problems. They analyze such data, trying to find patterns, trends, and drivers behind the customers’ behaviour.
This is a job where you’ll be the closest to the academic definition of science. It’s a research-based job that focuses on computing and programming principles and the resulting issues. By doing research, they try to improve those principles, improve current algorithms and programming languages, even invent completely new computing principles and programming languages.
As an ML engineer, you’ll build ML models and deploy them to production. This, of course, includes data manipulation and analysis. Based on that, you’ll build models and train them, as well as automate them by creating algorithms and AI software.
If you’re a data architect, you’ll be working on the company’s data management infrastructure and its architecture. You’ll have to understand the company’s processes and find technical solutions to support and improve its processes. This involves providing architecture from the point data enters the company in various forms, storing it into databases and data warehouses, changing it, and being the company’s output.
This is a job that puts you in charge of administering databases. That involves ensuring databases security, integrity, access to it, doing backups, restoring data. Also, you will be involved in database design and implementation.
As a data modeler, you’ll be responsible for data modeling or creating conceptual, logical, and physical database models. You’ll also be involved in database implementation.
Being a quantitative analyst means you’ll mainly be working with financial data. You’ll use it to build models which help your company make decisions regarding risk management, investments, trading on stock and FX markets, etc.
Your focus as a software engineer will be designing, developing, testing, and maintaining software based on the data and infrastructure provided by your other data science colleagues.
Statisticians’ main field of work is a statistical analysis of data. That means analyzing data with the purpose of finding patterns in it that will help the company make business decisions and predict the outcome of that decisions. They will also present those insights through easy-to-understand visualizations.
How to Get Started?
As you already know, there’s no universally accepted way of starting and managing your career in data science. However, a good start would be to get a degree in Data Science, Statistics, Mathematics, or any other quantitative field we mentioned when we talked about qualifications and skills.
After formal education, people generally start as data analysts. From there, they choose their path according to the opportunities, their abilities, and interests. Generally, you can go in the direction of working with data infrastructure and being a data provider, so to say. Those are jobs such as database administrator or data engineer. The other (very general!) path is being a data user, such as a BI developer or statistician.
As you gain experience and knowledge, you could be aiming to become a data scientist, ML engineer, or research scientist. Those are usually more senior positions, and they are, accordingly, among the highest paying jobs in data science.
Of course, you could also be coming from other fields of study and do these jobs. There are also some jobs that more explicitly require a little extra business, financial, or some other knowledge. For example, the quantitative analyst has to be very strong in finance. Business analysts, even though they have to be very IT-technology savvy, maybe it’s even more important that they understand the business than the computing side. Because of that, it’s not unusual that they come with a business degree but with IT experience. Or vice versa, a computer science degree with an extensive business background. It’s similar to the market scientist. They could be coming from humanistic fields such as sociology, psychology, philosophy, etc.
Finally, you’ll have to pass the interviews to get your data science career started in the job market. You’re wondering what questions you could expect?
For sure, lots of coding questions, such as this one from Amazon:
“Find the best selling item for each month (no need to separate months by year) where the biggest total invoice was paid. The best-selling item is calculated using the formula (unitprice * quantity). Output the description of the item along with the amount paid.”
(SELECT date_part('month', invoicedate) AS MONTH,
sum(unitprice * quantity) AS total_paid,
rank() OVER (PARTITION BY date_part('month', invoicedate)
ORDER BY sum(unitprice * quantity) DESC) AS rnk
GROUP BY MONTH,
WHERE rnk = 1
Statistics are important skills for data science, so don’t be surprised if you get a question like the one asked by State Farm:
“What is a p-value? Would your interpretation of p-value change if you had a different (much bigger, 3 mil records for ex.) data set?”
The above jobs breakdown and data science career advice should serve only as a guide. There are no ultimate truths that will make you a successful data scientist.
But what you can do to improve your chances of success is getting an education in the field we’ve mentioned above. Prepare for your interviews by going through the sets of coding and other questions. While you can’t cover everything for sure, being prepared for the types of questions is very important. After you get the job, it’ll hopefully be a job where you can use your knowledge and also learn more. You’ll see for yourself that experience is an extremely efficient way to learn and improve.
Then you’ll hopefully be able to choose more freely in which career direction you want to go.