In the era where data is the most valuable asset for a company, nurturing data skills has to become the topmost priority for any aspiring to mid-level data scientist.
The feeling of self-satisfaction with your current skillsets can land you at a miserable spot where your company may choose a candidate who not only has better experience in using new analytics tools but also has a deeper understanding of the latest trends. This, in turn, can result as an end of your career.
To solve this problem, we asked multiple data scientists from the AIM Expert Network community, to share their key insight on how to avoid unusual pitfalls and get out of the cocoon to build a bulletproof career.
From defining the right focus to analyzing the latest trends, here are the 6 pitfalls that data scientists should avoid to continually grow their career:-
Not defining the problem accurately
Most data scientists would ignore the “problem” in itself – assuming that this is a business responsibility – to discover or define the “problem” statement clearly. As a manager or a senior/lead data scientist, this is frequently a responsibility that should be within your purview.
Asking the right question is a lot more important than choosing the right algorithm, more often than not. This is a skill often overlooked by budding data scientists. And as we grow in an organization, when you need to scope out the larger data science project, we often stumble in this first step. Hence, it is very important for us to get into the “Questioning the status-quo” mode rather than get carried away by the details of algorithmic efficiency, the goodness of fit parameters or K-cross validations.
Sachin Dev, Senior data strategy manager – Accenture
Not nurturing the data
Data Science is made of 2 words “Data” and “Science”. If you take Data out of this then only science is left and that is what maximum folks are doing. It sounds like taking life away from Data Science itself. People forget to treat data as their child. You need to first spend time with your data, explore the data, get insights from the data and understand the data; later build ML/AI model as required.
If correct insights are not drawn from the data then how can one understand whether the model is behaving correctly? Unknowingly all focus shifts to just building the models and predicting the results. But that is not correct, 80% of work is done if data is treated to its minute details and insights are formulated. So be a Data Scientist and not just a Scientist who has worked on some data.
Netali Agrawal, Technology Lead – Infosys
Focusing only on coding
In my opinion “focusing only on coding” can end a data scientist’s career. Basically, data science is not about using python, R or Tableau; it is about understanding data and being able to make sense out of the data. Coding, as we are already seeing, is becoming automated with the help of tools that are able to generate ML codes. There are certain tools that are capable of building an algorithm based on the list which is provided by its developer. Just by adding a few inputs to the list, the tool will automatically generate the required code and execute it.
A data scientist needs to focus on understanding the data, how it is being processed and how it can be modeled to make better decisions. While technical/programming knowledge could be a good starting point and give one a relatively easy entry in this area but to sustain yourself, you need to invest in the concepts of data integrations, data modeling and data quality along with the understanding of business domain which will help you grow in the area of data science.
Just like the models that a data scientist builds have to be trained and improved continually, similarly, to be relevant and thrive in your career, you need to inculcate a habit of continuous learning.
Amit Agarwal, Senior Manager (IT) – Nvidia Graphics
Ignoring Visualization at the cost of Modeling
Visualization is one of the key skills, a data scientist must possess. Decision-makers & leadership teams will be interested in getting quality output in a visual format rather than just a code piece. It is essential for a data scientist to come up with informative graphs/charts explaining the essence of the work done.
Speaking with the Business team helps understand the visualization that can deliver maximum information to end-users. The output must be scalable enough for them to play around and make quality decisions to improve around. Customer Experience is more important than customer support as it ensures the journey is smooth enough.
Vijayakeerthi Jayakumar, Data Scientist – Cognizant Technology Solutions
Extreme focus on tools and technologies instead of fundamental concepts.
I have often seen people holding on to tools rather than the fundamental concepts. This is detrimental to any career, not to mention Data Science. Once you are addicted to a tool or a technique, stagnation is at the horizon.
A classic example amongst Data Science/ML beginners is biased (should be controlled anyway in ML) towards an algorithm over others. However, this is much more evident amongst potential Data Engineers, who prefer sticking to their favorite tools for ETL (Extract, Transform, Load). We need to understand that solving the problem at hand is much more important than using a particular tool.
Prasad Kulkarni – Senior Software Engineer
Disregarding the latest trends
Silently ignoring the latest news and trends in the world can lead to a decline in a data scientist’s career. The world is constantly changing. Data Science is all about problem-solving regardless of the domain. For example, you are working in an FMCG company, and are unaware of instances like economic slowdowns. You have a task in hand to predict the category forecasts for the next year.
If you are a sluggish person who is not much into reading news or the latest trends and working on a data model disregarding the external factors i.e. the latest trends. Your predictions will not be aligned with the current trends. This work may also not be appreciated by your company. The same example could also be applied to the ignorance of the latest technologies that come up on a regular basis and the subsequent upskilling ignorant patterns. Repeated cycles of ignorance of the latest trends can be harmful.
Ekta Shah, Data Scientist – General Mills
Avoiding participation in external and community events
We know that in today’s world, technological changes are constantly increasing, therefore it is necessary to update our knowledge regularly in order to prevent being obsolete. Especially when it comes to data science, we can see a fresh set of algorithms every month with their expanding spheres. This can be done in the following ways:
a) Reading a lot of research literature.
b) Going through a lot of magazines related to AI and its developments.
c) Participating in Meetups.
d) Trying to solve problems in websites like Machine Hack and Kaggle which keeps updating many contests. Many contributors also contribute different ideas of coding during the process.
e) Revising knowledge and trying to share our knowledge in websites such as stack exchange etc.,
Prasidha Ramnathan, Clinical research associate – Navitas Life sciences