What makes it hard to take the first step
Data science has experienced a glorious growth and popularity in recent years. Businesses are aware of the potential value of data more than ever. They invest in data science by means of resources and workforce, expecting a return in terms of business value.
The popularity of data science has driven a large number of people to decide to make a career change and work as a data scientist. However, it is definitely not easy to take the first step into the field.
In this article, I will elaborate on what I think makes it so hard to get your first job as a data scientist. Please be noted that this is my opinion based on two years of careful observation. Your opinion, feedback, or criticism are more than welcome.
Data science can be applied to any process in which we can collect or obtain raw data. What is expected from a data scientist is to create value out of raw data. The value can be in the form of improving a process, predicting unusual behavior, demand forecasting, translation between languages, and so on.
The ultimate goal is the same: create value using data. However, the techniques applied to raw data can be considerably different depending on the process or area. The tools and frameworks also vary depending on the given task.
For instance, a fintech company is likely to deal with time series data. Thus, they want a data scientist who is specialized in time series analysis. If you are to create a chatbot for customer service department of a bank, you must have advanced natural language processing (NLP) skills.
There are many areas that require to have some kind of specific skills on top of the general data science knowledge. This is where aspiring data scientists miss the train.
The problem is the lack of expertise in a specific area.
We try to learn as much as possible on a wide range of topics. As a result, we obtain a general knowledge on time series analysis, natural language processing, anomaly detection, machine learning algorithms, data visualization, and some other fields. However, we do not become an expert in any of them.
If you want all, you are likely to get nothing.
Data science is an interdisciplinary field with a broad spectrum of applications. It is hard to comprehend even the fundamentals of data science. On top of that, if we try to learn about each and every specialty in data science ecosystem, we are likely to end up being an ordinary candidate. Thus, we need to choose an area to master in order to outperform other candidates.
It is not only about theoretical knowledge. The tools and frameworks also are considerably different for different sub fields. Although the available tools make the job of a data scientist easier, it takes time and practice to learn how to use them.
For instance, if you plan to do natural language processing (NLP), you should be comfortable with Hugging Face. For a position focused on image recognition or classification, the foremost skills are the deep learning algorithms and frameworks.
If a big portion of your job is to deliver results or create dashboards, you should have decent expertise on a BI tool such as Tableau or Power BI. The list just goes on. The point is that you cannot be an expert on all these tools and topics. However, your chance of landing a job substantially increases with the level of expertise.
I think the optimal way is to decide on a sub field or specific area to focus. After obtaining a general understanding of data science and the basic tools, one should focus solely on a particular area. This is how you become an expert and increase your chance of getting a job in data science ecosystem.
The distinction of sub fields are becoming more clear. For instance, I have come across many distilled positions such as NLP engineer and machine learning engineer. If not the title of the position, the description almost always points out the desired specialties.
I know it is hard to decide on a specific area to focus. Moreover, you may not want to decrease the number of positions you can apply by focusing on only one area. However, you will stand out by demonstrating an expert level of knowledge. I think it is much better than being an average candidate for all positions.