As a data scientist you can choose working environments such as startups, corporates or research. Each type of organization has its own advantages and disadvantages. What are the differences and what impact does an organization have on your role as a data scientist? In this article Mylan Heijink (startup), Romain Huet (corporate) and Oscar Juarez Urizar (research) share their experience and vision.

Startup

Working at a startup can be chaotic, you ‘live’ for the next few weeks and are often unsure about what you’ll be doing in the months afte. You have to iterate a lot and take into account that plans can be changed easily. Mylan: “After working hard on a project, it can happen we have to decide to turn things completely around. You have to be able to deal with unpredictability. But when working at a startup you can contribute immediately and have a lot of impact. Our tech people quickly sit at the table with those who determine the course and are working on a complete module (and not just a small part). You quickly get a lot of responsibility and freedom. Developing a full solution solely by yourself is not an exception. Which is great, if you are cut out for this.”

Within a startup you’ll be working in a small team on several projects and solutions. Therefore, it is not surprising startups are more likely to hire someone quite versatile. Mylan: “Where startups benefit from is having developers who are able to fulfill multiple roles within a team. Depending on the size of your team and the budget you most likely have to do a bit of everything. Which I see as an advantage. This contributes in you solving issues and becoming resourceful. Startups have a fast pace in, for example, decision making. As a result, we work faster than corporate companies where everyone has their own specialism or works on their 'own' part within a project.”

Corporate

Corporate companies have many resources and plenty of money. They often organize knowledge sessions with peers and insightful discussions with colleagues from diverse background and cultures. Romain: “When working for corporate industries you’re able to develop a solution for a problem. However, things can take up much time which can slow a project down.”

The text continues below the info block

Within a corporate you can be working on algorithms development, dashboarding, stakeholder management, infrastructure maintenance, research, develop proof of concepts and product development. Romain: “My tasks and responsibilities are quite diverse and currently I am working on computer vision such as object detection, segmentation, classification. Also, I worked with natural language processing, tabular data (analytics) and timeseries (forecasting). Besides this, I align with stakeholders, maintain code base and infrastructure and am trying to find new use cases to work on. I think I had the luck to do many things which might not always the case for someone working in the same corporation for years.”

Research

Working within a research environment is exploratory and mainly interested in horizontal perspective, for multiple business. There is more direction to what you want to do compared to a startup, but less restrictions than working for corporate organizations. Oscar: “We aim to innovate and for that you need space, time and budget to explore. We don’t always know the outcome of a project. Sometimes you can share learnings immediately, other times you can use research in future projects. At the moment I am working on conversational AI and how to use new methods to improve customer support. To stay up to date every two weeks experts share their knowledge and experience about a specific topic. For example, one of the latest insights and updates were about virtual systems in Japan.”

Conclusion

As a motivated developer you are interested in data and want to get a lot of intellectual challenges and stimulation for real world problems. Choosing a type of organization to work for as a data scientist is about what you’d like to do and contribute. As a data scientist you’ll find differences within responsibility and diversity (and switching between) in tasks, but also in duration of projects. The most important questions to ask yourself are: Do I want to find solutions for every problem I encounter and am I able to withstand uncertainty? You’ll probably do great in a startup-environment. Or: Do I prefer long term, predictable plans and knowing what to do in my work? Then you’ll probably fit best within a corporate business. Another question to ask yourself: Do I want more freedom in exploration for research to be of value? And do I want to create a better solution or product which always asks for more information? Then research is your place to be. Most probably, the million-dollar-question to ask yourself is: which work properties will suit me as a person best? Something you can only find out by experiencing it yourself.