Each Friday we post a new v i s t a profile, a career beyond the academy story (use the tags at the bottom of the post to find the entire list). These posts accompany our curated events to support post-PhD career transitions, v i s t a mentoring, and also #sheffvista on Twitter.
Job title and company: Data Scientist and Community Manager at Pivigo
Approximate salary range for your type of role: Data Scientist (in London) starting salary is £35-45K.
Hello World! My name is Deepak, I am a data scientist and community manager at Pivigo, the data science marketplace. I started off my journey at the University of Sheffield where I studied Mathematics and Astronomy for my undergraduate degree, and then went on to gain my PhD in Astrophysics at Keele University where I was researching exoplanet atmospheres. At Pivigo I’m very fortunate that my role allows me to help other aspiring data scientists transition from academia, and I do this through university talks, webinars and mentoring on the Science to Data Science bootcamp (@S2DS) run by Pivigo.
How I moved from academia to the world of data science
Towards the end of my PhD, I thought “what is my next step?” and also during this time a good friend told me about data science, and the more I looked into it, the more I understood that it was a career path that was fascinating, would use my scientific skill set and gave me the opportunity to make an instant impact.
It became clear to me that this was a career I was looking for, so I applied for the Science to Data Science bootcamp and was accepted. I worked on a real data science project where I was part of a team of four building a recommender system for an IT company that provides equipment to education, healthcare and business. Not only did I gain vital commercial experience (in just 5 weeks!), but I also had great fun and met loads of really cool people. I’m also now part of the alumni network that is currently 550 strong and growing.
I went on to work as an insight analyst for a few months before my current role became available at Pivigo. Now, I continue to support other scientists looking to enter the world of data science as well as working on internal data science problems myself.
What I learned from this experience
When I was first looking to make the transition from academia into data science, I thought that the only sector I could go into was finance. How wrong I was! Data science touches every sector, and the projects you will be working on will be very diverse and exciting. The tools you will learn from one sector can be used across other areas too, and this means that moving roles is very straightforward when you are looking for the right position.
Furthermore, you may not work on a stereotypical project for your sector. For example, just because you have a job in a logistics company does not mean you will only be working on route optimisations, you could be working on understanding your customers through segmentation, predicting which of your clients may leave (churn rate) and identifying the reasons for their going, etc. You will also be able to attend conferences where you will grow your knowledge of the cutting edge of data science and see how other businesses are using data science.
As a data scientist, you can be a superhero, from optimising business to saving lives and everything in between. Over my nearly two years here I have learned a lot about what skills are needed to be a good data scientist and what I hope to share with you is the top 3 skills that you need to get started in data science NOW!
Skill/superpower 1: Computer programming
Computer programming forms the foundations of everything one does through any data science project. The two most used programming languages within data science are python and R. Both these languages are versatile and are relatively straightforward to learn either as a first programming language or as a new one. You don’t need to learn both, choose the one that you prefer. I would also stress that as with any skill, learning the theory (or in this case syntax) is not enough. You need to practice these skills as much as possible, whether this is for a project for studies/work or a personal project you are working on.
Skill/superpower 2: Machine learning
Machine learning is a big buzzword at the moment, and it might seem scary at first, but it is actually not as daunting as it may seem. Machine learning is a way of allowing computers to learn from data without being explicitly programmed. Supervised and unsupervised learning are two main types of machine learning. Supervised learning uses labelled data to make a prediction. For example, say you have information about houses in your area such as the size, number of bedrooms, is there a garden, size of the garden, etc. and the price of the house. You can use supervised learning to use the information about the house (features) to predict the price (target).
Unsupervised learning uses unlabelled data to understand patterns or trends in the data. For example, imagine you are a data scientist at Amazon, and the company wants to send targeted offers to its customers for Christmas. You can use clustering to see different groups of people based on features such as, for example, average basket price, purchases in each department etc. This is called customer segmentation. From there you are able to send the most relevant offers to each group. If you’re interested in some relevant reading, I have some tips and reading recommendations at the end for you.
Skill/superpower 3: Communication
One of the most overlooked skills one needs to be a data scientist is communication. Even if you are the best programmer and are able to work magic with machine learning, if you cannot communicate to your stakeholders your results and why you have done it then your work may not get implemented.
It is essential to learn how to present your work to both a technical and non-technical audience. The best way to upskill your communication skills during your academic work is to give seminars to your peers to practice your technical communication, and to general public audiences to practice communicating technical concepts to non-technical audiences. As an example, I volunteered at the Keele Observatory to learn how to express some of the awesome astronomy that has been done using the equipment and research Keele had done to a public audience.
If you can get a good foundation in these three skills, you will be able to be a fantastic data scientist. Also, continual professional development (courses, seminars, webinars, workshops, and self directed learning) is critical. Many companies will have the budget to help you upskill, and hence progression is swift. Also, the work-life balance is outstanding. Occasionally you may attend a weekend conference but the majority of the time you will not need to work evenings or weekends.
I hope you have found this blog useful and it has given you an idea of the areas that you need to work on. I am available to help with any questions you have, my email address is firstname.lastname@example.org. Please do not hesitate to get in touch. Pivigo also have loads of resources and practice challenges on our website so simply register for free to get access at pivigo.com. Data science really is the sexiest job of the 21st Century, trust me, I’m a doctor!
Tips and references
Machine Learning: Two excellent books to help learn about machine learning are Introduction to Statistical Learning with applications in R (ISLR) and Hands-on Machine Learning with Scikit-Learn & TensorFlow. As Uncle Ben said to Peter Parker “With great power come great responsibility”. It is important not to use machine learning as a pure black box, these two books will help you to understand how to use, and importantly, how to interpret your models. Both books help with practical examples, and with some practice, you will be ready to take on some really cool projects.
Where can researchers look for jobs like yours? Almost all businesses are using data science so any jobs board or company website you search will help you to see who is looking now. LinkedIn is also a great place to see who is hiring and what they are working on.
What professional/accrediting bodies or qualifications are relevant to where you work? At least a masters degree is required, PhD is a great addition but not mandatory.