It has been approximately one year since I decided to make a career switch from Civil Engineering to the Data Science. After working as a Data Analyst at Slalom for 3 months, I think now would be a good time to share my experience. I will try to present this blog as 3 distinct parts: Why – why did I make the decision to leave Civil; What – what did I need to consider before jumping into another industry; How – how did I make the shift as smoothly and successfully as possible.
Why – why do I leave Civil?
I took my undergraduate study of Civil Engineering at University of Toronto back in 2012. And in 2016, I chose to pursue my master’s degree in Civil for another 2 years specialized in Building Science (walls, roofs, and foundations stuff). After working as an engineer for about a year, I started to question myself if it is the career path that I want to take. When I talked with my university friends about my concern, it seemed that almost everyone was not satisfied with their current work and was seeking for an alternative path. However, most people are reluctant to go out of their comfort zone. More importantly, we don’t like the idea of letting go all the sunk cost (years of study, network, experience) that we have been investing for all those years and starting all over again. As of me, there are two solid reasons that push me to make a move.
One main reason is that the industry attitude towards innovation is still conservative. We see things changing every day, and everything around us is becoming (or at least trying to) more and more convenient. However, it is almost the opposite in the realm of Civil Engineering. Although there are many innovative ideas and products available on the shelf, it requires many different parties (developers, designers, contractors, administrations, governments etc.) to agree on adopting a new technology in the industry. It would be almost impossible to identify who should be responsible for a failure occurred 10 years down the road. Therefore, practitioners in Civil would prefer to do things in the old fashion only which has been proved working for the last 50 years rather than an innovation with unknown risk.
The second reason is that to gain professional experience in Civil is slow. Experience is the most essential factor for a person to be successful in almost any industry. In Civil, probably the only way to gain experience is through projects. Some may argue that there are many great articles, papers and books that you could benefit a lot. It is true in a sense when you are at a senior level and want to become an expert in a certain area. However, for junior folks, even with all the knowledge stuffed by 4-year university education, you could still look like a fool and know nothing on the site. Field/hands-on experience outweigh theoretical knowledge in the early stage of someone’s career in Civil. However, in Civil, you have to be patient enough to go through the entire project cycle (usually takes 6 months to years) and wait for the next project.
If that already sounds not too bad for you, life could be even worse with the other reason. I have heard so many stories about how people get pigeonholed to a specific type of work from my colleagues. If you have done a certain type of project (let’s say replacing windows), it would be more efficient to put you on the same project again from the company’s perspective. Eventually, you would become a worker on only a few streamlines. It is like a loop, and hopefully you could break by the chances like short on stuff or a very kind supervisor who cares your development more than company revenue.
Therefore, those two main reasons forced me to switch my career focus from Civil. For those who are having a mind-battling, my personal suggestion would be spend some time to figure out the true reasons behind the scene. You could get frustrated if the project is going nowhere or your boss is a jerk, but it does not necessarily mean that you need to leave your current industry and head into a brand-new field where the same thing could occur as well. Knowing what truly bothers you is one big step towards the solution.
What – what did I consider before jumping into Data Science
The next question I was facing was which industry I should pursue now. After filtering by the two aforementioned pain points, there are still many options available which are innovation driven and relatively easy to gain experience, such as data, finance, marketing, or even start my own business. “Choose a career that you enjoy and be good at” I know it sounds cliché, but it would be most sincere advice I can provide. Because every job could suck, like I said before, your daily projects could suck, your supervisor could suck, your peers could suck, and even the whole company could suck. Nothing is in our control, and it could be a real torture to be stuck in that environment for 8 hours every day. A high salary may ease the pain, but the chance to get paid really well is low if you don’t like anything around you. So, doing something that you are passionate about, not just for money or reputation, could be a better way to make those hours less miserable.
In general, I like logic and things that are self-explanatory. It really bothers me when I was in the previous company and asked a question, most likely the answer I would get was “it is just how it works, and we have been doing it for the past decade with no problem”. On the contrary, each single line of code would have its own purpose and there must be a reason for that piece of code to be placed. It is a pure joy when I first plan out the code structure in my head and start to build pieces until they work exactly in the way as designed.
In terms of things I am good at, probably the biggest strength I have would be my learning ability. Thanks to U of T, I realized my ability of learning the course materials within two weeks, making connection to the practical questions, and applying that to the final exam, which get sharpened each term. That might be exactly what your parents would not expect you to do, but it turns out very important for coding. There are generations of great minds who came up with those brilliant algorithms, and numerous hours have been spent to optimize the performance, so why to reinvent the wheel. All we need to do is to understand how it works (or not), and make it work in a data science project.
Of course, there could be other factors like job opportunity, salary expectation, and career development, but after doing my research and chatting with industry professionals, I feel that Data Science might be a good fit and worth to give it a try.
How – how to make the shift successfully
The biggest debate I have was that: should I learn everything online or sign up for a course to speed up the process. Although nowadays knowledge sharing is very common and you could find almost any information online for free, it could be overwhelming and inefficient to learn everything by yourself. Furthermore, there are more options in learning: data related master program, bootcamps, or online courses. Master programs would take the longest (1-2 years) and cost a big fortune for international students like me. On the other end of the spectrum, online courses could be cheapest and flexible in time, but it would also be hard to get personal questions answered and develop a comprehensive understanding of the field. Personally, joining a bootcamp seems to be the best approach to learn all the essential tools and knowledge with reasonable cost (3 months, $10k – 15k CAD). There are organizations that provides full-time bootcamp training like WeCloudData, BrainStation, and even UofT. One of my friends took the bootcamp hosted by WeCouldData before and recommended the same data science course to me, who by the way studied Chemical Engineering and now is the Data Scientist at Nestle Canada. Word of mouth means everything! I am glad that I took his suggestions, and I will explain in detail.
The first merit in WCD is the course enrollment policy. I signed up the bootcamp in early March, which will start in June. As someone with absolutely no background in coding, WCD allowed me to take advantage of their part-time courses to learn all the programming tools at no additional cost beforehand. During the 3 months, my weekly schedule was like this: taking Python and SQL courses during the weekends, full-time work on weekdays, and doing coding assignments/projects after hours. By the beginning of bootcamp, I was already very confident with my coding skills and had finished my first web scraping project which is an essential skill for any Python users. If you do not have any coding experience, no matter whether you are going to take, a master’s degree or a bootcamp, I would strongly suggest you working on the coding skills first. Do not procrastinate and think you would be taught during the courses. Watching at someone else coding is totally a different story than doing it yourself. It takes lots of time to practise and digest the syntax and coding logic, and trust me, you don’t want to be stuck on a single function while others are working on much complex projects.
The second point I appreciate is the course content. I have compared a few bootcamp syllabus offered by different institutions, and they vary a lot. Take the bootcamp I chose as an example; it is a 12-week program designed to develop a good understanding of the data pipeline end-to-end. It started with a review of Python, SQL, Tableau which are the three main tools if you are going to be working as a Data Analyst. Then in the machine learning section, I learnt a wide range of algorithms that are currently used by banks, social medias, and retail companies. It is also where most Data Scientists work on, building models and making predictions based on the data they acquired. For the last few weeks, it was on big data and cloud computing, which are the hottest buzz words that every company is trying to adopt, and lots of data engineer jobs falling into this category. Up-to-date is critical here in the data industry! Technologies invented 10 years could be considered as outdated, and definitely you don’t want to spend quite amount of time learning stuff that will be abandoned in next few years. Also, having an exposure of all type of work in data industry is extremely beneficial. I was aiming at Data Scientist before the bootcamp because it sounds fancy and making prediction sounds interesting. But later on, as I know more about the type of work they do, I tended to develop myself into Data Engineer who focus more on the data pipeline system.
Last but not least, the thing I benefit the most from the bootcamp is the real client projects. WCD has formed sustainable long-term partnership with many industry players. I have the chance to work on real data projects and built up my first work experience. Companies prefer to hire the candidates with some relevant experience, but how would a rookie get his first ever experience. I believe every career switcher have to face that kind of dilemma, and it bothers me the same as well. Since my technical skills were sufficient enough prior to the bootcamp start, I have spent almost half of time on the client project. The client I was working with is an e-commerce company, through which I have worked on a range of projects: marking campaign analysis, customer segmentation using machine learning, and database/data pipeline integration work. Unlike others who are also new to the industry, I would be able to showcase my skills on resume and make good impressions through talking about those projects during the interview. I think that is one of the major reasons for me to land my first job offer even before the bootcamp ends, and eventually to my dream job.
I have to admit that taking a bootcamp is not an all-in-one solution. There is no way that you could learn every possible detail about data science and achieve the same knowledge level as those who have been studying this subject for years, but I learn so much more than just coding in the bootcamp, such as integrating different tools, systematically approaching to a data problem, and ways to avoid data leakage. Those parts surely come with rich experience, which is pretty difficult to learn for a starter on his own. So, I would say the WeCloudData’s data science bootcamp is a really good place to start your career in data industry and provide the best tools for you to succeed.
Above concludes the entire story of how I make my career switch from Civil to Data industry. And, I hope my experience would be beneficial to you or at least help you to make the right decision that works for you. Good luck to anyone who is undergo the career switching process, and may you take your courage in both hands.
To find out more about the WeCloudData’s data science bootcamp Eric has taken click here to see the learning path.