5 Lessons I Wish I’d Known When I Started Learning Data Science

If you’re trying to break in, watch out for these eye-opening gems.

5 Lessons I Wish I’d Known When I Started Learning Data Science

One of my oldest friends called me up for advice, having started learning data science. He wanted to switch his field. I ended up having a lengthy conversation with him about my experience around what worked for me.

It was evident he was overwhelmed by the ton of resources available online and wanted to hear trusted information from someone who had successfully broken into data science.

While I was truly humbled, he chose to ask me for advice; I realised there could be many who don’t have a source of reliable information and could benefit from what I told him.

This article is for you if you’re starting to learn data science and in a similar position as my friend. In this article, I outline everything I wish I knew and what every lesson means to you for a better learning experience.

You don’t need a Master’s degree to break into the field.

There’s no way to be 100% sure unless you experience it yourself. But if you chose to look for examples in the world, and talk to people who are already in the field, they will tell you the same.

Many are already killing it in data science and artificial intelligence without relevant formal education, let alone a Master’s degree. I was lucky enough to be mentored by a Google Engineer, and his words gave me perspective.


When I started learning data science in 2017, I believed I needed a Master’s degree despite finishing up a computer science undergraduate degree. I recall outlining a plan for “MS in the US” and preparing for GREs.

The more I researched the popular master's programs and the modules they offered, the more it became clear to me. I can learn what they teach during the master's program, directly from them, at the comfort of my home for a fraction of the cost.

Wait, did I say that?

What this means to you:

Yes, I did. You can learn the same courses, from the professors from MOOC platforms such as Coursera, Udacity, and EdX for a much cheaper investment.

Which courses? Which platforms? I recommend you look at a previous article where I outline the resources I've used and recommend. I’ll summarise the alternative to Masters degree for you here:

  • Identify your strengths and weaknesses. If you’re coming from a different field, see how you can use your existing skills for your advantage.
  • Outline what you still need to learn and focus solely on them.
  • Rely on online courses from reputed universities to learn the concepts. I’ll be more than happy to help you with recommended courses based on your existing skills.
  • Apply those skills to create portfolio projects that demonstrate your expertise. The best way to solidify your learning is by applying it in a project.

It sounds simple, I know, but it works.

You don’t need a Master’s degree to break into the field, but you still need to showcase your expertise and skills.

You need not pick between Python vs R.

The first decision most enthusiasts need to make. If you had this question, I understand where you’re coming from. My short answer: you’ll eventually use both.

Coming from a computer science background, I was naturally comfortable with Python. I never liked R, mainly because I couldn’t get hold of the syntax, and Python was my go-to for most of my work.

Few months into my learning journey, I was lucky enough to bag an academic based internship under a professor. It was a data analytics internship and gave me the taste of the world of data science. There was one problem, though.

The professor I had to work with was comfortable with R. He asked me to use the same. If this happens to you, you’ve got two options:

  • Say no — and risk being one of the poor performers of the internship.
  • Say yes — but having to learn a new language from scratch.

I did the latter. I wanted to break into the field at any cost, if that means I’d have to learn a new language, let it be. I was thankful for this experience because I ended up adopting a learning mindset early on my journey.

A year later, for my first job, I had to use both Python and R. Now, during my current job, I primarily use PySpark. It’s similar to Python, but there’s more to learn. Thanks to my previous experiences, I overcame the fear of learning new languages.

I love being challenged, eventually to come out victorious.

What this means to you:

It’s often overwhelming when you’re starting, you would be required to make a lot of choices, and you’ll want to make the best one.

But don’t get into the debate of which language/tool is best. Why? All languages have a purpose in this ecosystem.

As a rule of thumb, if you’re coming from a statistics/mathematics background, I recommend starting with R. If you’re coming from computer science or other backgrounds, feel free to start with Python.

In the end, the language doesn’t matter, how good you’re with concepts of data science and artificial intelligence matters the most. The goal eventually is to become technology-agnostic; to use any tool/language that best fits the problem.

Your job title doesn’t matter; the kind of work does.

Most beginners (including me) go behind the title of “data scientist”. After all, it’s been branded as the sexiest job of the 21st century.

My first exposure to data science was as a data analyst intern. Before recently taking up a senior data scientist's role, I’ve also worked as a machine learning engineer. Strangely enough, I enjoyed every single role of I’ve worked so far.

I realized I was more focused on the work I was doing than the title. The questions I asked myself:

  • Am I learning something new every day?
  • Does the work I do have a real-world impact?
  • Have I gained the exposure to an end-to-end machine learning pipeline?

If you haven’t heard of the various roles in data science, data analyst, data engineer, business intelligence analyst, machine learning engineer, product analyst, business analyst, MLOps engineer and the list goes on.

All roles can be equally rewarding if you look beyond the title and start focusing on the kind of work you do.

What to look for:

As a beginner, please keep an open mind. The data scientist is an umbrella term, and the title doesn’t matter, but the kind of work does. You can be called a data scientist while doing mundane SQL querying, or be called a data analyst yet work on cutting edge technologies. Both exist.

I’d recommend to take up any role as far as it challenges you to learn something new. Continuously ask yourself a set of questions based on what matters to you the most, to keep a reality check.

Going behind titles will limit your chances of breaking into the field, and you wouldn’t want to do that. Trying out different roles will help you identify your strength. Once you’ve figured out what you’re good at, you may focus and specialize on it.

Approach startups for your first data science job.

The learning curve is huge if you are in the right place.

I got lucky with this one. To give you some context, I live in a small island called Sri Lanka, and there are only a handful of companies in data science and artificial intelligence. The job opportunities compared to the US or UK are slim here.

Since we don’t have many big tech corporates with data science jobs here, I found myself joining an AI startup, and I couldn’t be happier.

You don’t really need a big tech firm behind your name at the beginning of your career. The learning curve is much more important. If you find the right startup that resonates with your values and culture, you’ll feel home.

Startups offer a rich learning experience, with plenty of opportunities to grow. If you have the right mindset for a Startup, nobody can stop you from flourishing in your career.

How to approach startups:

Startups are all about people and culture. The best way to find out if you’d be a great fit is through networking with their employees. Here are some steps you could take:

  • Connect with the employees on LinkedIn and follow their social media account. This would give you a taste of their culture and what they work on.
  • Based on your understanding of the startups, shortlist the startups that you’re genuinely interested in.
  • Work on the required skills and build demonstrable projects to showcase.
  • Reach out to someone from the team, through an email or over LinkedIn, informing that you’re interested in the work they do.
  • If there’s a vacancy, it will eventually work out. Be patient.

This is a real story of a few young passionate data enthusiasts who joined the AI startup I was a part of. They did exactly what I listed above, and we were more than happy to welcome them to our team.

It’s never too early to start sharing whatever little you know.

Someone (and your future self) will benefit and appreciate it.

When I started learning data science, I felt like I had no knowledge to share. I mean, who would possibly want to know about my experiences? I’m not an expert by any means. I was convinced to write by Rachel Thomas of Fast.ai and finally took the step to share my knowledge here in a blog.

Two significant changes happened:

  1. Many people could relate to my experiences and started benefitting from my work. I constantly received positive feedback and was more than happy to help everyone who reached out.
  2. I started understanding the concepts I wrote, much better. To learn is one thing, but when I tried to explain it in an article, I had to do further research and synthesize my learnings and explanations. This practice enhanced my understanding of the core concepts.

I have started benefitting more than how much my readers have benefitted. Learning something new and sharing it with my readers has become a habit now. I wish I started sharing much much earlier.

“To teach is to learn twice.” — Joseph Joubert

How to start sharing:

When you’re progressing through your learning journey, indeed, you don’t have much to share. You still can start by sharing your experiences on what you learned, what worked and what didn’t.

This could be on your blog, Medium, or any other platform such as YouTube. Think of it as teaching a friend to learn with you. You will easily find like-minded individuals from these platforms.

The more you learn, the more there is left to learn. As a beginner, you don’t want to come out as someone with bragging and know-it-all attitude. Always remember there’s a lot to achieve even if you feel like you have already achieved significant.

Be humble and patient with all kinds of feedback you receive when you put yourself out there. You’ll never know what’s in store if you don’t start.

Final Thoughts

Thank you for reading so far. I genuinely hope this was of use to you, especially if you’re starting to learn data science. In this article, I summarised the lessons I shared with my friend over an hour of conversation:

  1. You don’t need a Master’s degree to break into the field.
  2. You need not pick between Python vs R.
  3. Your job title doesn’t matter; the kind of work does.
  4. Approach startups for your first data science job.
  5. It’s never too early to start sharing whatever little you know.

I extensively write about my learnings and experiences in the world of data science here in Medium. The community was helpful when I started, and this is my way of giving back.

If you have a better idea of how you can approach the world of data science and feel confident you too could break-in, this article has served its purpose.

Did it?

For more helpful insights on breaking into data science, interesting collaborations and mentorships consider joining my private list of email friends.