Tips for Data Science Enthusiasts: From an Engineering Degree to a Senior Data Scientist

On the value of engineering and problem-solving skills to stand out in data science, Giulio Giaconi from Italy is currently a senior data scientist at Ofcom. He has 5+ years of experience in both academic and industrial research. He worked as a teaching assistant at Imperial College, London, where he completed an award-winning Ph.D. thesis. After successfully completing an Omdena real-world project on applying machine learning to improve road safety, we asked him about his learnings and tips for fellow data science enthusiasts.

Can you describe your journey into becoming a senior data scientist? 

I graduated from Sapienza Università di Roma, “Telecommunications Engineering” 2011, and I went on to pursue a Ph.D. in information and communication theory, Imperial College London, 2018. I became interested in data science and machine learning during my Ph.D., partly because these fields flourished and gained a lot of interest in the academic community at that time. For this reason, my first job after my Ph.D. involved researching machine learning techniques for cybersecurity at BT Applied Research. From then, I moved on to my current role as a senior data scientist at Ofcom.

The problem-solving mindset I acquired during my engineering studies was crucial in helping me transition into data science. I believe that to be a data scientist, it is essential to learn how to formulate a problem in the right way, understand and formalize its hypotheses, and set out a plan to address it, which are all skills that can be developed with any engineering, scientific or numerate education.

The best advice I got a while ago that I want to share is:

“Take your time exploring and fully understanding your data before using any machine learning technique or developing any model on it. Although this may be seen as a boring activity, this is in fact a crucial way to understand the problem as much as possible and get some vital intuitions about how to solve it.”

Can you share a point in your career where things got a bit difficult? And how did you overcome roadblocks? 

One of the most challenging points during my career occurred perhaps at the very beginning of my transition into my first data science job. I was juggling two tasks: wrapping up my Ph.D. research and dissertation on information theory while trying to learn as much as possible about a technical field, machine learning, which I had only briefly touched on before. Back then, I remember splitting my time very rigorously between the two tasks, and my working days ended up being really long. What helped me was setting out a very clear and feasible schedule for both tasks to avoid losing focus and motivation.

More generally, when I experience roadblocks and difficulties, I tackle them with an open mind and eagerness to learn. Also, if a problem is technical, it is easy these days to find a great deal of accessible information online – this is true, especially for the machine learning and data science fields.

I also want to stress that there is a large number of capable professionals that can help you with either technical problems or even with questions on career and who can be reached online, whether it is on social media or directly on more technical sites like StackOverflow or GitHub. I truly recommend reaching out to people you think may help with technology but also more general career-related questions – you may be surprised by how many people are willing to help if you reach out with a genuine question and an open mind. 

How did the Omdena experience help you? What tips can you give to current Omdena collaborators?

Omdena allowed me to work on a topical global challenge, preventing road accidents worldwide, by using a variety of machine learning techniques. 

I appreciated that I could use my skills on a project tackling such a crucial problem. Although I was already familiar with most of the techniques used during the project, I still gained a number of insights, thanks to the contributions of very skillful collaborators and the considerable amount of information continuously shared.

Moreover, what I particularly liked about Omdena was the collaborative and social dimensions of the project. I really enjoyed the fact that from the very beginning I felt part of a community because of the many enthusiastic and supportive peers I was surrounded by, all of us set to solve one challenge together. I think this is a pretty distinctive feature of Omdena compared to other ways to learn and practice data science skills. This opportunity helped me test and also improve my soft skills, e.g. communication, organizational, and leadership skills.

My main advice to current and future Omdena collaborators is to engage as much as possible during the weekly meetings, be inquisitive, and offer constructive feedback and advice to each other. While working on the project, you should strive to create a safe space where everyone feels free to give their opinions, and you should challenge others fairly but also expect to be challenged at the same time.

Any closing words?

Now it’s an exciting time to start building a career in data science, and there are countless opportunities for people who want to strive in this domain. There are many ways to learn about data science and machine learning, including books, online courses, challenges, and tutorials. More importantly, I would definitely recommend getting your hands dirty on a real problem using actual data. For this reason, I strongly encourage working on an Omdena challenge, not only to quickly get up to speed with data science and machine learning, but also, and more importantly, to have the chance to solve key challenges affecting millions of people globally.