Practical Experience is what matters in Data Science??
I have been planning to write this blog for a long time now, there is always a question that gets raised whenever an “Aspiring Data Scientist” starts his or her journey in this field, the question is “How to get started?” or “How to land my first job in Data Science??”
Let me share with you my own story as I was once struggling to answer the same question, and of course, the path I took may not be helpful to all my fellow commoners and I am aware of the fact that there is always a bit of luck component where you are beginning your journey and who are your peers or the managers when you are beginning, but I am optimistic it will give you a sense that the title of this blog is a real question :).
Chapter 1: Who was I?
Graduate of NIT Jaipur (2008–2012), Chemical branch, started my professional career as a Java developer, being from a Non-IT education stream and starting your journey in Software always poses up some different set of challenges, but I won’t be boring you up with them and won’t be telling you the company names I have worked for as a developer. The most important part of telling my history is that in 2015, I came to Eko India(https://eko.in/).
Chapter 2: How did it all start?
What we hear most people saying related to the Data Science field is that the love for Data and Maths is imperative to start your career and I could not agree with them more but according to me the practicality is that you have to love the business where you are working more than any other thing to become a Data Analyst/Scientist, you have to understand the ins and outs of the problems what the business or that particular sector is facing. All other things whether it's data or maths tend to fall in the right place when you start loving the business.
And this is what drove me towards data science the most, being in Eko makes you fall in love with the business as the company was a growing startup at that time in the fintech field and the most important thing was that the company was open to new ideas and the founders were up to take new challenges and lay the foundation of things which can help them grow in the future.
I won’t lie by saying that the charm of being called a data scientist did not contribute to my aspirations but still the enthusiasm of making things work or introducing new things in the company inspired me more.
The problems which the company was facing or for that matter what any company face, there is a role which data plays in that and that is what made me start looking at the data which the company was having. Though most of the data were internal and transactional data, for the start it was enough to lay the foundation of the beginning I was looking for. Though I admit it was not easy as I was a developer and at that time the company did not have any “Data Science” team to which I can switch and be a part. I started doing what we call exploratory data analysis and started handling some of the data requirements of the Finance department which usually comes for a PPT to be presented in an investor meeting, but that gave me a lot of sense about data. For ex. in one of the requirements, I need to present a cohort analysis of all our retailers and that simple cohort analysis gave everyone a sense of the problem which of course was in the back of the mind of everyone but was never proven through data.
Now the company as a whole was looking to start the data journey together and not just me, I was given the opportunity to work on some of the ad-hoc outsourced company-driven projects like location analytics or setting up a data lake/warehouse internally with the help of external vendors. This does not provide to me any hands-on experience as such but it surely gave me a chance to work with the external people who were a part of this field and it got me in touch with every department to understand their problem to be data-driven in their day to day work which is again for me a most important part of becoming a Data Scientist.
This also helped company to correct the missing data which was not transactional but was more related to the profile of their customers, which is one of the starting problems for any company on how to make their data more and more consumable and truthful.
Till now I have not taken any formal education for this, all the analysis was done in simple python notebooks, MySQL, and excel, but now it was time for me to take a step ahead and learn some things which can only be learned by studying, i.e the theoretical part of being in this field, but all this experience of 1 to 1.5 year gave me most important learning.
There are problems which a company face can be sorted out without the use of Machine Learning Algorithms, the major task is to identify those problems and to solve them in a simple way
Chapter 3: The Process
As the company was growing and now it was looking to ramp up the process of setting up a Data Team in-house and solve more problems with the help of data, I also started growing my knowledge database by taking up a few courses. I started in 2017 with the first batch of Upgrad(https://www.upgrad.com/) with the course https://www.upgrad.com/data-science-pgd-iiitb, this is a wonderful course if you want to start your journey, it starts with the basics and more importantly, it gives you a practical depth to understand how a problem can be understood from a data science way along with the theoretical concepts which are anyways important.
I also started some of the data science interviews along with the course, though not for changing but to have an experience. I failed a few and you will find one of my interview experiences below.
I also cracked one of them and that too without any major practical experience in the Machine learning field, and this give you a hint about the answer to the question asked in the blog title :)
Obviously, I did not take that job as the real journey for me as a Data Scientist in this company has begun just now. Below are the things we did as a company.
Setting up a Data Team
Setting up a Data Lake (Vendor based)
Power BI dashboard for reports ( With BI expert in the team)
Picking up Machine learning Usecases( Clustering Retailer universe, Solving Churn problem, KYC Images classification). I will be discussing these use cases when I write a blog about the respective algorithms, my next one will be the Random forest which will cover the Churn use case practically along with the theory of the algorithm.
Convince businesses to utilize our data solutions, reports (The toughest task for any data science team)
This all happened with the beginning of my course to almost 2.5 years, but it was a great learning for me, the team, and for the company as well.
Chapter 4: Who am I now??
Taking a practical experience for almost 5 years in the same company I decided to move on and take up new challenges to grow myself. I moved on to one of the biggest Data Science consulting companies and scaled up a POC project over there, from there on I am right now in one of the biggest F&B companies as a Senior Data Scientist handling the Data science vertical of the Data Analytics team over there.
Chapter 5: Learnings??
Coming back to the question Practical Experience is what matters in Data Science??, here are my personal two views over this.
- Yes, it does to grow in this field, what more important is to understand the business/market you are working in.
- No, it is not necessary to land your first job in the Data Analytics field, you can develop the business sense and take a course, to begin with, and start giving the interviews.
Gratitudes
- To Eko(https://eko.in/) as a company and all the people over there
- To Upgrad(https://www.upgrad.com/) teachers and mentors for helping me to lay a foundation.
