Become a Data Scientist or die tryin’


Cancel your netflix account, now it is time to start your data science career. 2 years ago I started as a Business Intelligence Analyst in the Data Platform Team at Stylight, but quite soon I noticed there is so much more than basic analytics in spreadsheets. I want to share some insights I got along the way to accelerate your data science skills.

1. Work in an encouraging, thrilling environment that motivates you to go beyond your daily tasks

Working hand in hand with experienced and skilled developers will help you a lot to get ramped up quickly. You get support when you are stuck and get tips for free all along the way, where you sometimes would spend hours of researching. Let me give you an example, the first time I tried to install a jupyter notebook, which is currently the preferred UI for many data scientists, my notebook just refused to let me install all the needed dependencies and I was stuck. Sergii, Data Scientist in our team saw my struggle and stayed until late night to get things sorted and I could get started. Situations like these can make the difference, so choose the environment you work in wisely.

2. Never stop learning – do as many online courses as possible

I am convinced that workplace learning and lifelong learning will gain importance substantially during the next decade. Online courses at one of the well-known platforms like coursera, udacity, edX and so on are an affordable and compelling way of acquiring state of the art knowledge. Nevertheless, there is a huge variety of courses you could possibly take, and a lot of new fields to explore. I would recommend to start with the basics like learning how to retrieve data with SQL and to get familiar with the command line and learn a solid programming language widely used in the data science world like Python. From there you can move on acquiring more in depth skills in statistics and machine learning.

As already said, there is exuberant wealth of possible places to start, but I made good experiences with:

Python for Everybody Specialization (University of Michigan on coursera)

This unpretentious course of Dr. Chuck offers a great way to get your hands dirty on some Python

Machine Learning Machine Learning (Stanford University on coursera)

Oldie but a goldie. Andrew Ng’s famous ML course is taught in octave, but there is no other course out there giving you such a dense learning experience to get started with machine learning.

Applied Data Science with Python Specialization (University of Michigan on coursera)

This is a brand new course series with to the point videos using jupyter notebooks for course assignments

Apart from online courses it is great to read up what you learned in books like:

  • Python Data Science Handbook by Jake VanderPlas
  • Introduction to Machine Learning with Python by Andreas C. Müller and Sarah Guid

Be aware that these books have git-repositories where you can find the code used to explore it more in depth.

3. Go to meetups and conferences

Socializing is always fun and at meetups people share their experiences, struggles and solutions for free. This is a great knowledge pool where you can get inspiration for new tools to use or approaches you might want to try out. Depending on your level of expertise you can start with talks that give you a high level overview of the data science field and then dig deeper to check out more specific talks, for example about the hottest tools around like Tensorflow, jupyterlab and whatever else makes you happy. I would especially recommend to check out the Pydata and PyCon events or in general Numfocus events. Apart from the tech-knowledge you will gain you can discover new cities all around the world – and I am just saying, there even is a PyCon Jamaica!

4. Find projects within your current environment to improve your skills and apply them on a real world problem

Learning online and talking with peers on meetups and conferences is one thing, but what really brings you up to speed is applying your learnings in your daily job. It is just the way your brain works, with daily repetition of patterns you will consolidate your learnings from short-term to long-term memory where – as a fun fact – your hippocampus plays an important role. Apart from that you will get to know your tools better, manage to look up things you don’t know more quickly and this will finally bring you to a whole new stage of enjoyment.

If you cannot find a project within your current job right away take part in some online challenges, for example at Google’s data science platform kaggle – to see how other people approach problems can give your learnings a huge boost.

5. Do I need a data science degree to start a career in data science?

Data science is an interdisciplinary field with many different facets. At universities data science programs generally consist of statistics, maths, computer science and more. I am sure this gives you a great foundation to do data science in practice. Nevertheless, it is part of the beauty of data science to attract like minded people from various backgrounds. I hope that also in future data science will be characterized by the variety of cross functional folks rather than limited to a number of domain specialists. Because no matter what your background is, the one thing we have in common is our passion about the insight we can get from transforming and visualizing data.

Along your way you will learn a lot of things you might not need right away, but don’t worry. It is pretty likely that you at least will need the concepts you learned in your later career. You will struggle and sometimes have the feeling that you start from scratch all over again. And every single day you will have the feeling that there is so much more to discover and understand.

But remember, no one is born a master and there is no free lunch. Is it still worth the effort? Hell yes.

You are passionate about data and want to work with us? We have an open position as Data Engineer in our team.


|By Stefan Tippelt – Business Analyst|


Image by beachmobjellies (Attribution-ShareAlike 2.0 Generic) at flickr.


Comments are closed.