Job hunting sometimes feels like looking for a candle in a dark room we’re unfamiliar with. We don't have complete information, every company has different standards, and it’s unclear where a position might take you in the coming years. This is true, especially in data science because data science work is far from being standardised. There are many layers and levels of how different your work might be than another data scientist’s.
Previously I talked about what type of positions you can end up in (in-house/consultant data scientist/freelancing/etc.) and the difference between the seniority of positions. But there is yet another layer to this: the team structure.
It might sound like a trivial thing to know but team structure can give you hints on what type of data science work will be expected of you and which direction your career will grow towards.
Understanding different team structures and how they affect your data science work will also make it easier for you to choose between companies. Not to mention that it will give you a solid reason for applying to places so when you get asked “Why our company?” you will have something of substance to share.
Some options for how you would work as a data scientist are:
Depending on these positions, you will have varying levels of responsibilities and creative workload.
[Not sure how AI and data science relate to each other? Check out the Data Science Kick-starter Mini-course]
AI or DS teams tend to be freer when it comes to their work. Their goal is to develop relevant tools and technologies for other teams in the company or for the end-user. It's not uncommon for companies that have AI teams to dedicate part of their effort into research. These are companies that understand the value of AI and what it can bring to their field of work and understand that creative or research-based work is necessary to keep up with what’s happening in the world and adapt latest technologies. That's why, in an AI team like this, you are more likely to do creative work, pursue new technologies and apply things that were very recently invented.
The work in these teams tends to be start-to-end. You are likely going to need to deliver complete products. Thus, a mix of skills might be needed. People in these teams tend to have the freedom to specialise and grow in an area of their choosing as long as it aligns with the company goals.
If you are one of the few data scientists in a department, your job will more likely be one-shot quick tasks such as creating an analysis, building a model on data. There is a chance that the work will be standardised and you’d be expected to do similar things periodically. It is not likely that you will be expected to build standalone state-of-the-art products. Teams like these tend to want their team members to produce reliable, low-risk solutions. You will be using proven, well-established techniques. There will likely be plenty of freedom to make your own choices as you will be one of the few people who know about data and how to work with it. As you gain experience in these teams, you will likely become an expert in a certain department (this could be marketing, sales, etc.) and have a career as a data expert in that area.
If you are the only data scientist in the team, it is even less likely that you will create products. You will be, though, the go-to person for anything to do about data. It is a dangerous place to be in, in terms of responsibilities as the expectations might be very high. But if you like taking responsibility and being in charge of things, it might be a blast. The work will likely be fast-paced and day-to-day as there will be a lot of input coming from different parts of the company and you would need to provide the data support needed. If the company is not a high tech or data-driven institution you might end up working with very old technologies that are being abandoned mostly by technologically more advanced firms. Though in a company like this, you are likely to be closer to business decisions. You, most likely, will have a higher effect on business decisions and you might end up being one of the most valuable assets of the company. Your future in this position will likely tend towards the business side.
And there you have it, these are some observations of how different teams use data scientists and how work differs based on team type. Don’t forget of course that these are merely observations and not rock-solid facts. Everything I talk about in this article is based on my own experiences and what I have heard from my podcast guests so far. You might find a company where you are the only data scientist but you are doing rockstar level research. Or you might end up in a data science dedicated team where machine learning has not been used yet. Still, knowing possible team structures will give you an idea of what you want and what you don’t. And you can ask more direct and targeted questions during your interviews.
So it’s time for you to decide, do you want to work with state-of-the-art tools, produce standalone products, build models in a fast-paced environment or help businesses grow faster? You don’t have to make a decision that is set in stone, of course, decisions change, dreams develop into other dreams. But at least having your goals defined in a brief sense will help you search more efficiently and will inspire confidence.