Blog
What does a career in data science really look like?
Like many things, transitioning from academia to commercial data science is, arguably, simple when you know how. But without insider knowledge of when, where and how to get a job in the data science industry, many STEM PhD graduates find themselves uncertain where to begin on their career path to becoming a data scientist.
Last week, I sat down with three data scientists who have successfully made the transition from academia:
• Kyle Jarvis is a Data Scientist at HSBC.
• Emma Cooke is a Machine Learning Engineer at Robin AI.
• Gokhan Çiflikli is a Senior Data Scientist at Attest.
We touched on multiple aspects of what it takes to join this exciting and sought-after career path. The consensus was a positive one: this is a really good time to be, and to become, a data scientist; the same ‘buzz and excitement’ that enticed our data scientists to make the change is even greater today.
So, what is it really like to be a data scientist today? We covered a broad range of topics during the event, but five themes really stood out to me.
Data scientists have the chance to make a real impact
The panellists noted that the impact of their work as a data scientist can be seen quickly. What you create can be deployed to solve today’s problems – not to solve problems from decades in the future, as is often the case with academic research projects. Gokhan argued that the best thing about being a data scientist is the constant intellectual stimulation involved with solving real-world problems, rather than just the theoretical.
Almost every industry needs data scientists
Emma highlighted that, when compared to academia, one of the main attractions of a career in data science is the diversity of industries available to you. From finance through to healthcare, data is a constant in every industry today; that means that data scientists have a huge and diverse range of career options open to them.
Commercial data scientists implement new techniques almost as fast as they’re discovered
Panellists noted that ‘data science’ is a new and ever-evolving field, and consequently it’s important to keep up to date with the latest techniques. Fitting this in alongside other work responsibilities can be challenging, but it’s also very exciting – the techniques you research today could have real business impact tomorrow. Other common challenges faced involve the amount of data you have to work with – whether you’re overwhelmed by vast amounts of unstructured or messy data, or struggling to piece together small scraps into a cohesive whole, the dataset is rarely perfect.
Working in data science provides plenty of opportunities to push boundaries. Kyle noted that, whilst a lot of data has been around for decades, there is plenty of untapped potential that data science and machine learning is just starting to uncover. Becoming a data scientist today means not only getting to use new techniques, but also sometimes being able to discover new applications of them too.
Day-to-day data science work is incredibly varied
Hearing our speakers discuss their current areas of focus made it very clear that ‘data science’ is a complex and varied discipline. Our three speakers are all involved in different kinds of data science projects at their companies, showing just how diverse data science is as a field. Projects range from using natural language processing (NLP) to mark up contracts and NDAs in a more time efficient way, through to using active learning to make the most of partially labelled data and help with fraud detection.
Succeeding in data science is often about finding a team that fits you
Anyone wanting to make the transition from academia to working as part of a data science team generally wants to know about the assessment process. The panellists said this is as much about checking whether you fit in with the team and the company culture as it is about testing your technical skills and knowledge.
Other challenges are dealing with non-technical colleagues’ perceptions of exactly what data science can achieve or solve. This means it’s crucial that anyone going into a data science job is able to communicate to non-technical colleagues, and of course interviewers, how machine learning and AI can be applied to real life business problems.
Industry programmes are invaluable to help PhD graduates kickstart their data science careers, with Emma citing Faculty’s Fellowship programme as one of the key things that helped her acquire the skills and industry experience she needed to move into data science.
If you’re interested in making the transition from academia to a career in data science, you can find out more and apply to our latest Faculty Fellowship programme here.
