At the end of my master in Computer Science, I knew pretty well what I wanted to do. Leverage my technical skills to solve business problems. To that end, I enrolled in a couple of management-related electives during my final CS years. After this glimpse, I decided to deepen my business knowledge, and I pursued an extra master, the Master in Management from Universidad Carlos III de Madrid.
New country, new school, new learning. Today, I am glad I went for it. On a professional side, I acquired knowledge about strategy, finance, marketing ..., but also many soft skills. Indeed, the teaching approach was centered around group work and presentations. This was a great training for me, it allowed to practice communication and team management skills, with students from different backgrounds and countries.
On a personal side, I discovered a new city, met new friends and I could cheer for my favorite soccer team, in their unique former stadium, the Vicente Calderon.
Kicking off my career
With a strong background in Computer Science and Management, I was ready to leverage my technical skills to solve business problems. Back then, my flatmate was working as a Data Scientist for a Big 4. He explained to me how he was helping TelCo companies to optimize their shop presence using their internal data and Machine Learning algorithms. What he was doing seemed awesome to me, and I went on reading about the Data Scientist roles, entitled as the sexiest role of the 21st century by the Harvard Business Review.
Now I just had to find a Data Scientist job! Easier said than done... Indeed, for such position, a company will grant a lot of responsibility to the employee, as he or she will have a strong impact on its business operation. It can lead to great efficiency improvement, or to a waste of time, depending on how competent is the Data Scientist. Therefore, firms will look for medior profiles, especially if their analytics team is small. Hence, I struggled to find a junior friendly position, which would give me the required experience, the chicken and egg problem.
Resiliently, I eventually discovered a great opportunity at a local boutique consulting company, specialized in Google Cloud. There, I worked on many interesting projects. From adjusting the marketing campaign of a leading car manufacturer thanks to machine learning generated customer segments, to increasing the understanding of a NPO’s fan base by developing an intelligent social media listener tracking an evolving list of keywords and hashtags built on natural language processing.
I also meet great colleagues, such as Charles, who became a cloud expert and recently launch his own consulting venture, 10/10 would recommend!
The McKinsey datathon
As I was cruising within my Google Cloud Data Scientist role, for about half a year, I received a LinkedIn message from a partner of McKinsey Belgium. He explained how they were creating an analytic hub in Belgium and that they were organizing a datathon, i.e. a Data Science oriented hackathon, to which I was invited.
Of course I accepted. At the time I was casually active on Kaggle, thus aware of the kind of contest I could expect. I was looking forward to being part of such an in-person, Kaggle like, competition. Moreover, during my Spanish stint, I had heard of the great place that McKinsey is, and how it could give me access to even better and unique opportunities.
Couple weeks later, I joined the event with another 100 participants and ... I won first prize!
I was so excited about the win that, on my way home, I punctured a tire of my car, because I was on the phone and not focused enough on the road - ahah.
On top of the cash prize, I was also invited to the interview process at McKinsey. It was a unique opportunity, as, in theory, they would not hire junior profiles to kickstart their Belgian hub. But since I demonstrated my abilities, they made an exception for me.
I pursued the recruitment procedure, and ended up getting an offer. Back then I wrote about this whole experience and it was shared in a blog post.
The interview
A couple of traces of this interview still remain online, a tweet and an Instagram post. On the other hand, the direct link to it is not active anymore.
The text below was originally published in October 2018.
Why did you decide to participate in McKinsey Analytic’s March hackathon?
It started when I received a message in my LinkedIn inbox. The mail was a solicitation to participate at a McKinsey Hackathon. It also said that this contest was taking place in the process of building the Advanced Analytics hub in the Benelux. The invitation was very appealing to me for two reasons:
- I was very curious to discover how McKinsey is handling the analytics competences;
- I love to work on Data Science problems, for me it is similar to solving a brainteaser or an enigma. Thus, I decided to apply for the event and, a few days later, received a confirmation.
What was the most fun and interesting part of the hackathon? (actual Data Science Challenge)
I enjoyed the context of the hackathon. We got a dataset containing records of patients and the goal was to predict their survivability. For each record, we had physical information about the patient such as age, sex, weight; as well as the disease and the treatment. I liked to work on this topic because it is very impactful. Indeed, the forecast could help doctors to give a better medication to the patient and eventually save lives.
What did the recruiting process look like after the hackathon?
After the event, I planned a call with a recruiter from McKinsey Advanced Analytics. During the exchange, I could ask my questions on being a Data Scientist at McKinsey and I received the information about the hiring process.
Then, I have been in contact with a consultant from the local office, in Brussels. We talked over the phone about the Advanced Analytics department and did a small simulation of a case interview. The consultant gave me feedback and advised me on how to prepare myself for the real interviews. The last step before the actual interview was a coaching session at McKinsey’s office. Other applicants and I met in person with a consultant. Together, we solved a full case and discussed about the personal experience part of the interview.
Finally, I had my first round of interviews a few weeks later. Each round consists of three interviews, with three different interviewers. Two of them were general business interviews and the last one was focused on analytics. Later the same day, I received positive feedback and the invitation to the second and last round of interviews. After two weeks, I went back to McKinsey’s office for the final round of interviews. The process is the same but you meet more senior people from the company.
Overall I enjoyed the recruitment process, the discussions I had with employees from McKinsey were interesting and it felt more like an exchange than a one-way interview.
Who do you recommend participate in this kind of hackathon?
I believe that anyone with a genuine interest in analytics is a good candidate for the hackathon. Indeed, you will have the opportunity to meet with fellow passionate people, test and strengthen your Data Science skills and finally, discover the careers that McKinsey can offer.
What did you enjoy the most about the McK Hackathon? (whole event)
During the event, I liked to discuss with employees from McKinsey. After the competition, we could connect with other participants and consultants from McKinsey. I enjoyed having an informal talk with them and being able to ask my questions on the practice. I learned about the different types of roles within the Advanced Analytics practice and how a typical project is handled.
How did you prepare for it?
My preparation was a day-to-day effort. Being a hired Data Scientist, I expected the hackathon to require similar competences as the one I practice in my daily activities. In order to put all the chances on my side, I focused my learning in three areas:
- I made sure to keep up to date with the latest algorithms in Data Science;
- I put an emphasis on being proficient with data wrangling and preparation;
- I went over a few Kaggle competitions to have an idea about the different types of challenges that exist in Data Science and to look for best practice.
What would you suggest to future McK hackathon participants?
For the hackathon I attended, efficiency was the key of my win. Indeed, the length of the challenge was limited to four hours. Thus, we could not afford to lose any time. My recommendation would then be to master your basics. One should be quick at identifying the type of algorithm one is going to use. For this, make sure to take into account the goal of the task (classification, clustering, …) and the type of data (image, tabular, …). Be also prepared to deal with some data engineering and cleaning. For example, if you decide to work with Python, as I do, be ready to wrangle data with Pandas.
The goal is for you to have the time to try different models and work on feature engineering.
Final thoughts
Eventually, I did join McKinsey as a Data Scientist, I was followed by two colleagues the next month. I went on and worked with many great people, over 10+ countries and the rest is history.
As a side note, I was not the first ever employee to work with data science at McKinsey Belgium, but I was the first Data Science Fellow, i.e. a client facing consultant, who leverage his knowledge in that field to implement AI models, from inception to production.