The Exact Steps I used to Become A Data Scientist @Microsoft

Briit
Total Data Science
Published in
8 min readJun 29, 2022

In this article, I share my personal experience of how I became a Data Scientist at Microsoft.

In my previous article, I laid down the steps you need to take in order to become a Full Stack Data Scientist. In there, I have given the steps you need to take step-by-step and I even pointed out that it is advisable to get a mentor, although not mandatory. In this article I will share my own experience with you during my learning journey, my job search stage, my interview stage and my work stage and most importantly give you my advice at the end. I believe this will motivate you and give you a clearer picture of what is involve in the journey of becoming a Data Scientist.

My Learning Journey

I did my first degree in Mathematics, back then in 2009 ‘Data Science’ was just a tiny word known by few folks. It wasn’t something attractive to go for as a Career compare to something like Software Engineering, Accounting, Law, etc. I worked as a research assistant and I was primarily involved in data collection and management for Accreditation Council for Business Schools and Programs(ACBSP), formerly the Association of Collegiate Business Schools and Programs.

ACBSP is a U.S. organization offering accreditation services to business programs focused on teaching and learning.

Working with ACBSP is when I started developing much interest in working with data. When I left the cooperate world to pursue my masters, very few schools actually have programs related to working with data, I mean as a profession. The hottest IT related career to go for was software engineering. I started my masters as an engineering student wanting to become a software engineer, however, my ultimate goal was to work with software tools and systems related to data management. I wasn’t getting much of data related concepts to learn during the time as it wasn’t even the priority of schools to include in their curriculum.

I started searching for random learning materials and videos on youtube relating to Database Management and SQL to get started with. Machine learning related task were mainly done by statisticians which they mainly do forecasting and less robust ML predictions. But that was the top notch skills you could find as of that time. There wasn’t much robust ML algorithms and systems to build any state of the art ML models. I reached out to most of these guys on Linkedin to know more about the field.

Somewhere in 2010, things began to shape up when participants in competitions such ImageNet came up with much robust techniques to perform Machine Learning.

Fields like Data Analysis, Business Analysis and so on grew so fast within a short time. The trigger was the influx of smart phones, iPhone was a game changer and most mobile producing companies started imitating the features in iPhone, even websites were built with smart phones in mind. Majority of online interactions were made with smart phones and that made it so easy for people to generate data and even offer their data for free unaware.

Smart companies like Amazon started collecting data about their customers, analyze and draw insights from it. Building recommendation systems became the order of the day for these companies. Platforms like Facebook, Youtube, Instagram, TikTok became the major beneficiaries.

However, to get the right skills to handle these data was the major problem. There was no proper structure or path to learn these skills as the field was very much new. Colleges had nothing in place for that new shift. No specializations for Data Science or Data Analytics like you see today. Even the so called top schools were sweating to carve curriculum for this field.

In 2010, very few colleges started Bsc and Msc in Data Science which was really had to get in and the cost was super crazy. The most you could find was a post-graduate certification course or a short certificate course. Most of these courses were in a bad shape and merely scratching the face unlike courses like The Full Stack Data Scientist BootCamp® which dives into details of the concept for a better understanding.

I managed to switch from core software engineer I enrolled for initially to data science. Although I enrolled in data science, I can tell you on authority that most of the things I got to master during my college days were the things I took responsibility to learn outside my traditional classroom. The likes of online courses, youtube videos, research papers, blogs, and things like that.

I needed to do the best for myself.

I later on put together the exact steps I took during my learning stage which I have laid down below for your reference.

  1. Start with SQL
  2. Master Python programming
  3. Master the necessary Statistics
  4. Master Machine Learning
  5. Have a general understanding of the various concepts in Artificial Intelligence (AI)
  6. Specialize either in Computer Vision(CV) OR
  7. Natural Language Processing(NLP)
  8. Master one Data Visualization tool (either Power BI or Tableau)
  9. Get familiar with cloud platforms like AWS, AZURE, GCP, etc.
  10. Get involve in Hackathons and Internships (paid or unpaid)
  11. Keep up to date with the research papers
  12. Keep your Github account up to date with relevant projects

The above sequential steps is what I took in becoming a Data Scientist at Microsoft.

You can find more details here

My Job Search Stage

Although the field of Data Science and Analytics looked promising, it was only a few companies that understood the importance of data that were hiring data scientist. Companies such as Amazon, Google, Facebook, Microsoft and a couple of them understood what they can do with their customers’ data. However, there were lots of other companies that did not even understand the field of data science let alone thinking of employing personals for that. So it was more like if you want to do something meaningful with your Data Science skills and also get some decent salary, then you need to make your way into one of these big companies and you know getting these companies is not easy, it means getting really on top of your game.

And oh, there was nothing like campus placement for Data Science. These big companies were going for PhD students with some considerable amount of work and domain experience. They were just scared of letting any amateur employee handle sensitive company data since they can be sued and that will bring hell to the company.

I first joined Microsoft as a customer service insights specialist…could you image that, someone who dream of becoming a Data Scientist is now holding this kind of position. Even getting that wasn’t on a silver platter. It involves dealing with the customer data and as I pointed out these companies were skeptical about letting you handle the data. I applied for several positions and those days, as pointed out earlier, if you do not have a PhD, getting a Data Science job was only a miracle unlike today that you can get it even without a degree(oh yes, two Data Scientist in my team, each receive $117,000, and do not even have a single diploma, just online course BUT they know their stuff well, which is all that matters).

Before joining Microsoft I had 3 job offers, one as a Database Manager with a startup, second as a Data Engineer at IBM, and of course as a customer service insights specialist at Microsoft. Then I chose Microsoft, weird I know!! But i had my reasons: 1) I bargained for a good salary at Microsoft using the other two positions as a leverage. 2) At Microsoft, you can easily climb the lather if you know your stuff, so I knew soon I will get the position as a Data Scientist internally once am in. 3) The work I was going to do as a Data Engineer at IBM was almost same as what I was going to do at Microsoft with even a little flexibility. You get my point?

Not getting exactly the position you are looking for initially can be disappointing but if you set long term goes instead of short term goals, at the end it will all come together to make sense.

My Interview Stage

The interview stage is probably the most scary phase of your journey. You do not know the one going to interview you and he/she does not know you either. He/she does not care how much effort you have put trying to understand the concepts, or how badly you need the job. It’s non of their business. It’s either you prove worthy of being hired or go to hell!!..sorry.

But if there is one thing I got easier, I will say is my interview stage. WHY?

I only have my projects that I have done so far present at my Github page so I included my GitHub link inside my resume. Recruiters want proof of concepts. Certificates do not really prove any point. After all anyone can get the certificate even if they have not mastered the concepts. These online course creators are desperate to give you the certificate so that they can get rid of you and enroll others.

Before I appeared before the interview committee, they had gone through my Github profile and knew what skills I have. The interview was all about the projects in my GitHub repositories and how I went about doing them. I was in control because they were asking questions about projects I have done and I could answer them with confidence because I did the projects and I understand them. There wasn’t random questions that will make me fumble just what I know. If I had no GiHub repository with my projects, they would have been in control and asking me questions that even my teachers cannot answer..lol.

The technical interview wasn’t much of a challenge either because….well maybe my GitHub profile had already convinced them to hire me. I was able to do most of them-not all though but most with confidence.

My Work

When I started work as a customer service insights specialist at Microsoft, I wasn’t really comfortable mentioning my position to friends and family, not that the position wasn’t good but I used to brag that I will be a data scientist right after college and here I am as a customer service insights specialist…lol

But being focused, I transitioned to my dream role in 6 months of my joining date. I used to work my ass off to impress my supervisor. Today I have absolutely no one to impress that am good. In fact others have to impress me because I can get them fired (oh yes I can!).

In a Nutshell

So you see, It’s all about knowing what you want and preparing well for it. In todays world there several resources to easily learn Data Science, at the same time there also more people interested in learning Data Science-meaning competition is also high.

My advice

Spend time to learn your craft well, don’t rush and you will thank yourself a million times later.

I have personally put together this course to help you easily learn Data Science and become good at it.

— — — — — — — -

You can connect with me on Linkedin for any enquiries.

If you like this article, kindly show appreciation by clapping for it. Thanks in advance.

Stay safe and never stop learning!!

--

--

Briit
Total Data Science

Data Science | Artificial Intelligence | Machine Learning