It was not until three years ago or so that I heard the phrase Data Science Unicorn which refers to a superhuman data expert who’s not only capable in digesting numbers — and magically mold a model out of it to classify, predict or detect whatever peculiar things you keep on your data cauldron — but also master the ability to write some spells that keep your data intact. Well, for the later part, let’s say how to transfer hundreds of thousands of clicks per second on your web page safely to a cluster-computing dungeon before transform it into a real number that would be likely useful. Put a wisdom to ask the right questions and answer the business goal on top of those ability, and voila, that is your so called data science unicorn.
I was in the middle of learning data science when I heard such a term and it honestly excite me. Well, how can I not be excited with the prospect of becoming one of the most beautiful mythical creature ever.
However, as mythical as its name, being a perfect data expert is close to impossible — at least for an ordinary human being like me, because there could always be people out there who is smart and dedicated enough to gain a such a super power.
As I left college and once again back to the industry, the excitement had turned into fear. I felt terrified on how people might expect me to be an almighty data warrior while all I knew was just tiny bit of this and that. It’s getting worse since I can’t decide on which part in the realm of data science that I want to explore more. However, setting all the worries aside, I moved on and joining a Data Analytic team in a company which stand behind some mostly visited media websites and one of influential messengers in Indonesia.
Being inside the circle and facing the reality on how the industry exploit data to support business growth, I can now actually see one way to put a partition on what to learn when one wants to learn about data.
The first group is the data engineers. Simply — yet brutally — put, programmers with fancier title. Or, if you prefer it, a software engineer whose niche is anything around data. Those who play with the latest big data technology offered out there and relied their life upon map-reduce; those who know that spark and storm has nothing to do with fire and weather while knowing what exactly hadoop is; and those who do not think of bee nor honey when heard the word hive.
The second partition, I call it the data scientists. These people are the reincarnation of what used to be called statisticians, who deal with numbers, familiar with bayesian inference and well understood the difference between mean and median at the least, obviously.
The third one is the data analysts. These people are the ones that actually discover informations from the works of the data engineers or even the data scientists. They are the ones who do complex queries, manipulating columns, aggregating rows, drawing charts and graphs that eventually bring knowledge leading to any decision making, business development or product improvement.
While in a sense, there is a clear line that could make the life of data scientists, engineers as well as analysts easier, the fact offered in the industry could never be ideal. It is true that these days, most of tech-aware companies would have a data analytic team, or anything equal to it, but what exactly these teams do may vary. One company may need more machine learning thingy while the others may just need someone who can select some rows and put the result in a single spreadsheet page. On the other hand, one data analytic team may also need to do all the things mentioned in the beginning of this page.
If you are someone who is started getting interested in learning data science and wondering what you should learn, it would be beneficial to take a step back and re-think what you really wanna do, and what your true strength is. You might be more interested in learning math than keeping up-to-date to latest technology and code-programming. You might also be more interested in honing your decision making skill and be in the front of business development. Or, if you want to take the challenge, you can always try to learn all the three aspects and be that unicorn you always wanna be.
ps. that company I work for? https://blog.kmkonline.co.id/