In this post, I will show my personal opinions regarding data roles. The emphasis on opinions is designed since the purpose is not to provide any kind of career advice but just to share some perceptions observed working as a computer scientist.
In 2017, I started my career as a data scientist only because it was the role that matched the things that I was studying in my master's degree course. I remember working with visualization, optimization, data analysis, ETL preparation, ML modeling training/serving, and also communicating directly with the customers. Maybe this was common to many people due to the broad job description of the role. Of course, it is not the optimal way to structure roles, but since it was a startup, you discover that everyone is accumulating responsibilities.
Apart from how chaotic this routine was sometimes, it opened my eyes to how the whole business of the company works and helped me to have a mental model of how the things that I was doing connected to the pieces of other engineers, business analysts & sales were doing. For me, this sense of how the business you have involved works, might value more than many years of technical learning - with all due respect, and I obviously think having both is the best - because it drives your focus/prioritization on things that will be more beneficial for this 'business gear'.
Note that I am not saying that you must prepare features for a risk model, automate a multi-instance deployment, create dashboards/reports for monitoring, and implement business rules to use the model predictions. Just be aware that these job descriptions vary a lot based on several things: project, team/company necessity, market, and last but not least, YOUR preferences. For example, my transition from data scientist to Machine Learning Engineer (MLE) was shaped by an off-the-job study of MLOps practices and their many benefits for companies, becoming a global trend in data. I started to incrementally bring up discussions about some development practices that we could adopt, highlighting the operational benefits involved. Fortunately, I had a lot of support for the ideas that I was bringing, but mainly because the value to the business in the mid/long term was clear to them. I don't get promoted or changed my role to MLE just to this initiative and that is the point: tech roles are natively liquid roles. Today, training a Neural Network for speech recognition is on your scope and tomorrow might be creating an API to accelerate partition distribution over many clusters could be your new scope.
My claim here is that adding business value to the company can extrapolate your job description and you could be aware of this by focusing on things you like to do/study/improve even if it goes beyond your current role. Tech roles come and go but at the end of the day what is important is if you are adding value to the company or not and making it clear to the stakeholders (of course).
Top comments (2)
Absolutely! If the company doesn't provide reasonable incentives it is very difficult to have this "ownership" mindset.