Rosaria Silipo, PhD, now VP of data science evangelism at KNIME, has spent 25+ years in applied AI, predictive analytics and machine learning at Siemens, Viseca, Nuance Communications, and private consulting. Sharing her practical experience in a broad range of industries and deployments, including IoT, customer intelligence, financial services, social media, and cybersecurity, Rosaria has authored 50+ technical publications, including her recent books: “Guide to Intelligent Data Science” (Springer) and “Codeless Deep Learning with KNIME” (Packt).
Recently, in an exclusive interview with Digital First Magazine, Rosaria shared her professional trajectory, roles and responsibilities as Head of Data Science Evangelism at KNIME, insights on the future of data science, personal hobbies and interest, future plans, pearls of wisdom, and much more. The following excerpts are taken from the interview.
Hi Rosaria. Please shed some light on your career path and how you started your career in data science and grew to your current role of Head of Data Science Evangelism at KNIME.
My career in data science started many years ago. I graduated with a master thesis on neural networks for the analysis of biomedical signals. Since then, I have worked often on data projects, either data engineering, data analysis or more properly data science projects, in a number of different industries, like healthcare, banking, finance, cybersecurity, marketing, and more. Multiply this for many years (my graduation took place in 1992) et voilà …I have built a quite large technical expertise in all that is data related. To that, add a pinch of passion for teaching and in general human interactions and I became a data science evangelist. Over the last year I built a whole group of data science evangelists, and I became the VP of data science evangelism at KNIME. Our group takes care of education and upskilling around data science and KNIME software – including instructor-led courses, self-paced courses, and certifications -, relationships with the education institutions, bits and pieces of data science educational content on social media – see our “Data Science Pronto!” series on the KNIME TV channel on YouTube -, and of course we also take care of the current large KNIME community.
Often, I am asked if I started this career conscious of my passion for data or just by chance. At the time of my master thesis, data science was not really a profession, yet. Some people were experts in some domains, had a thing for numbers and data, and somewhat ended up in the data path. I cannot really say that I consciously chose this path. I did like numbers, and I did like to find out what those numbers meant. So, yes, it was a choice dictated by my passion for math and numbers. However, how much of a career plan I had in mind at the time I really cannot say.
What does the future of Data Science look like?
A question that I hear more and more often these days is: “Is data science dead?”. I mean, with the new rampant AI engines, is it still necessary to train models? It is a legitimate question. In my opinion, with the introduction of AI, the approach to data science has become much easier. So, in that sense more people will adopt and consume data science-based solutions. However, we will still need expert implementers, not only to build AI engines, but also to extract insights from company data, build dashboards, and especially store large, clean, and usable amount of data to feed self-developed or AI based data science applications. We will have more generalist data scientists, leveraging what AI has to offer, and less, but still present, more specialized data scientists of the “old” school.
What types of career opportunities are available in the field of data science? Could you please share some insights?
Among the many things that my job entails there is also hiring. Usually, companies search for three categories of data professionals: data analysts, data engineers, and data scientists. In my group, since we work so much with education, we also look for teachers of data techniques.
In general, a data engineer’s job is to make the data available to applications, algorithms, and users in the form and shape that is needed. This means, data engineers are responsible for collecting the data over time and from multiple data sources, then storing it, cleaning it, transforming it, and finally making it available to external users and applications. All of that must be done according to the country’s law and the company’s governance, efficiently, and correctly.
A data analyst on the opposite is responsible for extracting information from the data and submitting it in front of the stakeholders in the form of reports, dashboards, and data apps. They are responsible for the creation, update, and maintenance of such reports, dashboards, and data apps.
A data scientist capitalizes on the data structure created by the data engineers and uses it to train, optimize, and deploy machine learning algorithms. The professional figure of the data scientist in the past was responsible for all data operations. But recently, other two specialized professional figures emerged: the data analysts and the data engineers.
Data engineers are surely the most requested profession on the job market at the moment. All of these AI models, with tons of parameters, must be trained and updated with myriads of data. Data engineers make sure that this is possible
With 10+ years of industry experience, according to you, what skills or characteristics make someone a seasoned data scientist?
Technical Experience, of course. There are two dimensions to becoming a seasoned data scientist, both related to experience. One is the extent of the data techniques and algorithms we have worked on; the other one is the depth of experience in one or maybe a few areas. For example, an experienced data engineer could be specialized in just a few databases or have the know-how to operate many databases at a superficial level. Also, a neural network specialized data scientist and a more generic data scientist using a variety of machine learning algorithms, both qualify for the title of seasoned data scientist, in my opinion.
Interpersonal skill is definitely another feature of a seasoned data scientist. It might seem that data science is an exclusively technical domain, but the capability of communicating results, explaining the importance of a project, and negotiating resources often decides your career in the long run. To progress in this career, you need as much technical experience as much communication skills. Producing the best results, but not being able to explain them to a less technical audience is a recipe for disaster.
How do you keep yourself skilled and relevant in terms of knowledge?
To keep up to date is indeed a challenge. You need to reserve time during the week, during the month, to read about and to apply hands-on the new techniques. A trick could be to get assigned to a project that uses them. But, as the head of the group, I might not always get what I want. So, I must make sure I still set some time aside to keep up to date.
To keep being relevant is easier. Though the techniques change, some features that make you an expert do not. Whatever is being built, it still needs to be efficient, scalable, fast, correct, … and this is true now as it was true then. The requirements do not change as much also for the new techniques.
In your academic or work career, were there any mentors who have helped you grow along the way? What’s the best piece of advice you have ever received?
There have been many mentors over the years, even now that I am definitely “seasoned”. Some mentors have been mainly technical, but the best piece of advice I got was on how to assemble a group of data professionals. Technical skills are important, of course, but sometimes when hiring you can compromise on seniority in exchange for a positive attitude. Many so-called geniuses (often self-proclaimed) are a nightmare to work with. On the other hand, sometimes you want to have that person in your group, maybe not even the smartest one, but who is able to keep the group together. These people are priceless when running a group. I learned that in a conversation with my ex-boss of twenty years ago. She used to be my boss, and we still meet from time to time to exchange precious advice on leadership and work.
Which technology are you investing in now to prepare for the future?
Of course, like everybody else out there, we are trying to increase our skills and experience in AI engines. It is not too complicated, but we need to keep up to date on the chances that the new technology offers.
Another parallel track involves verticals. We need to teach ourselves but also our users about new technologies. Our users are often highly specialized professionals in some data field. Thus, at the moment, we are researching new techniques for users working in finance.
What are your passions outside of work?
I am a quiet person, you know. One of the things I like the most is to make my own olive oil. I have a little garden in the middle of nowhere with 7 small trees. Every November I make my 8-10 liters of olive oil. I collect the olives, bring them to the mill, and then collect the oil. It is incredibly satisfying. The best part is to collect the olives from the trees. You are outside for a few days, relatively nice weather, moderate physical activity, quiet, no internet, no emails, no slack messages, lots of time to think. And then of course the smell of the fresh olive oil.
Where would you like to be in the next 5 years?
The company I work for, KNIME, has a bunch of offices around the world. In the next 5 years I would like to open a new office in Italy (we do not have that yet) and then become the head of the Italian branch of KNIME. The KNIME community in Italy is one of the most vibrant communities we have. It is a lot of fun to take part in the Data Connect events (the events organized locally by KNIME community members) in Bologna or in Palermo. It would be really cool to run an office there!
What is the one piece of advice that you can share with other professionals in your industry?
For all data scientists, remain curious! The technical skills in our profession change quickly. You need a curious attitude to remain up to date with the least effort.
For the heads of a data science group, hire smart collaborative people! A group, whose members work well together, make for half your job and allows you to dedicate more time to more interesting tasks.