Follow Danny on his self-taught data science journey
If you want to get into data science but you’re not sure how, or you’re not confident about it, try following Danny Ma on LinkedIn. His down-to-earth, experience-sharing memes will reassure you that you are not the only one feeling that way.
Read on to find out more about this jovial data scientist on his self-taught data science journey.
Huifang: What does it mean to be a self-taught data scientist and ML engineer?
Danny: I feel that almost every data scientist working in the industry is self-taught to some extent because there is no real curriculum that tells you: this is how you do data science. For instance, I did my undergrad in Actuarial science. It’s like business statistics and insurance models, kind of STEM but not really at the same time.
In my own journey to becoming one, I can honestly say that I learned almost everything on the job.
When I needed to dig deeper into a subject, I did more research using Google, Stack Overflow, reading documentation or books.
In recent times, there are degrees that are trying to bridge the gap between the education that is required to teach someone to become a data scientist. Even if one comes with a STEM degree, there is still a big gap between what we are taught at university and what we actually do in the workplace as a data scientist or machine learning engineer. Because this area is so new, no-one really knows what is best practice.
For me, a self-taught data scientist and machine learning engineer is someone who is very curious with a persistent problem-solving approach. Don’t give up immediately when you run into a problem. Instead, keep trying with workarounds from different angles.
Huifang: What makes you want to become a data scientist?
I had felt stagnant in my career after a year and a half of doing SQL work and data problem-solving. In 2011, I read in the Harvard Business Review that data science is the “sexiest job of the 21st century”.
I didn’t even know what data science was back then and I had to Google it. After Googling it, I stumbled upon all the data science stuff that resonates with what I had wanted to do with data in the first place, which was just solving difficult problems.
Huifang: What is the toughest part about being self-taught?
For a period of time, I woke up at 5:00 am every morning to study Python for three hours. I read machine learning blogs and I tried winning in Kaggle, in an attempt to gain recognition as a data scientist.
As I really enjoy learning new things, it was more of a challenge than a chore. However, I was still a data analyst with zero data science experience.
This was difficult for me when I started going for interviews. I was always asked if I have ever done machine learning at work. I would always answer no – and then never hear back from them.
Huifang: Transitioning from a data analyst to a data scientist, when do you feel confident or good enough to call yourself a data scientist?
I was interviewed for a data analytics role in a bank. I accepted the role, but I told my hiring manager that I really wanted to become a data scientist. Eventually, I was given an opportunity to work with the data science team on a project to help the business make more money. It was at that point that I was officially offered a change in job title – from data analyst to data scientist.
I felt like an imposter as I didn’t have enough experience or a PhD. However, this feeling dissipated once I started working with really strong data scientists who mentored me. I learnt a lot by working intensively on a large project for six months. Since then, I haven’t really struggled with any projects.
The guys whom I thought were geniuses were just regular people like you and me. They just enjoy solving problems and could do it faster.
They were happy to share their knowledge about what worked and what didn’t. This way, I wouldn’t make the same mistakes they did.
Huifang: What’s your advice for people who want to follow in your footsteps?
If you are in some sort of analyst role and you want to move into data science without much experience, try to move into a team that doubled up data science with an analytics function. You will get to work with real data scientists with whom they can ask for advice and build a good relationship. This might take a little bit longer, but it’s a natural progression and a good strategy to get the help to move into the area.
Alternatively, try your luck by applying for different things or connecting with different people who might just have the open-mindedness to take a risk on someone who doesn’t have the experience, but has lots of passion.
Huifang: We can easily make a switch from our existing role to a data science role if we are at the right place, at the right time, within a business. However, it’s not necessarily the same for the younger generation where a matching qualification is required during hiring specifically for a data science position. What do you think of this?
I considered myself really lucky because that’s exactly what happened. I had the right mindset, the right boss and the right people who could support me in a place with a lot of data. There was less competition as we got in relatively early. Our employers would prefer that we upskill in certain technologies or techniques to try solving business problems. Whilst now, I feel that hiring managers may be biased towards hiring new people with the corresponding expertise instead. Existing employees should focus on what they are doing well on.
It’s not really a problem as there is always going to be a market for people who are good problem solvers. Machine learning, data science and data analytics techniques are just part of the toolset that we would use to solve problems.
The most difficult thing would be proving to a prospective employer that you are a good problem solver.
Follow us for the second part of the interview to find out more about how Danny built an engaged data science community from scratch, fulfilling yet another career aspiration.