My journey to data science was a fortunate stroke of serendipity. Immediately after my graduation, I joined Robi as a specialist in the Technology division. However, my role was reasonably far from what data science entails, and more along the line of network engineering. I wasn’t happy at work and hence, 6 months after joining, I started to apply elsewhere. I applied to GP for a role in the business intelligence team without knowing much about what awaits me. From there on, I never looked back twice.

No matter how big an advocate I am of data-driven decision making, I have made a lot of important life decisions on a whim. My risk appetite is quite high and I enjoy traveling off-the-beaten paths (yeah, literally as well. I am an avid trekker). Now looking back, I realize my decision to pursue a business school PhD in the US was influenced more by the nature of uncertainty and excitement (I only knew a handful of Buetians who went along that road) than by my true love for the academics. It only took a semester to realize what mistake I had made. But I enjoyed the coursework and decided to continue until the masters degree. Coming back to Bangladesh was an easy decision. I went abroad not to settle but to have an experience.

While I was in the US during 2016-2018, I closely watched how the startup ecosystem in Bangladesh was evolving. I really wanted to be a part of it, and this willingness made my decision of ‘What’s Next’ easier. I knew a relatively well-trodden path; Apply for OPT, Look for Jobs in the US, Apply for H1B and then settle in the US forever. Instead, I decided to take a leap of faith. I had a few competing offers when I came back. One from a well-known MNC, One from a successful startup and another from an egg that barely hatched. I took the last one. I joined as COO and lead the product, design and tech. The startup eventually didn’t take off and I left for Intelligent Machines.

I’ll be honest here. If we strictly talk about the courseworks, then my bachelors degree (EEE in BUET) didn’t help much but my masters degree (Economics at Purdue) helped relatively more. But, data science is more than just the technical understanding of how to run a machine learning model. Soft-skills matter. If someone is trying to portray a different picture, he is probably lying.

The field is new and a melting pot of scientists and practitioners from many disciplines. So, there is no lingua franca as of now. A Statistician's take of data science would be significantly different from a Biomed or CS grad. A lot of people try to establish the superiority of certain domains over others which I find pointless. I take data science as a tool and not as a discipline. There is another unique problem/challenge I face which is specific to our local market. So far, decision makers at large MNCs and local conglomerates don’t have sufficient understanding of how AI works. They think that it’s a panacea that will cure all the problems. They also think AI is something very precious or fragile that has to be handled with care. Unsurprisingly, AI has become more household than electricity and the earlier we realize it, the better.

Everything that I have read, watched, and experienced contributed to this. There isn’t one single book that you can complete and become a data scientist. It’s more of a journey than a destination. If you want to have a genuine impact in this field, a solid academic foundation is essential. If you want to be more on the applied side, do not deprioritize domain knowledge over technology.

The issues are twofold to be honest. Firstly, the companies often don’t know the right size of the team required to solve an analytics problem. Right-sized teams are rare. Right size not only means the quantity of personnel, but also implies the right mix of talents. Secondly, retaining talent is difficult due to migration. After the pandemic, the talent crisis will probably be even severe. A qualified developer today has more options than ever before. Remote jobs are aplenty.

What I consider my biggest success is building a great team. I am really proud of the team members I have. Believing in diversity and letting it reflect on your work are two different things. I have been successful in building a team that is diverse, has a great sense of ownership and delivers results. When you have a great team motivated to achieve greater heights, amazing results are just a by-product. I am lucky to have a long list of such by-products in the last couple of years. On the other hand, however, I realize that a vibrant community of data science practitioners is missing. Now, with so many companies and startups working on such interesting ideas, an open and transparent community that shares knowledge and promotes AI ethics would propel this industry significantly forward. I feel partly responsible for this as a member of this community that I haven't taken any initiative till date to make that happen. This is also probably what I consider my biggest failure.

If you are an undergrad, focus on your studies. Grades do matter. Try a couple of interesting projects during semester breaks. Participate in a Kaggle competition. Read business journals. Meet new people and explore new domains. Learn with genuine openness. Avoid any forum that preaches Messi is better than Ronaldo or something similar. On the flip side, if you are a career-switcher, think twice why do you want to become a data scientist? Avoid the gimmicks of ‘becoming a data scientist in 7 days’ and ask yourself if you are passionate enough to grind for hours more than your peers if needed.

I can’t think of any particular book that single-handedly impacted my journey. But, I used to read a lot about Macroeconomics and politics during my undergrad. I was more curious to explore the underlying mechanism of ‘why things happen’. Data science has been a great career in that regard.