Several people, particularly from academia, have approached me asking how to get started with a career in data science. Obviously there are many online resources on this topic, nevertheless, here is my take on this. Feel free to send feedback if this was helpful or if you have something to add.
To start off, I'd like to explore what makes a good data scientist? Of course, there are plenty of blogs on that one [1-3].
In my opinion, essential skills include: good problem solving skills (finding a solution, even if it is not perfect), project management skills (being able to see a project through end to end), excellent communication with business/non-technical people, and being able to work well in a team. The latter might have been acquired during an M.Sc., most certainly during a PhD. However, some people have these skills naturally and might have proven so in other positions than a degree. In case you'd like to work on some of your soft skills have a look at this list of book recommendations 
Regarding the basic technical skills (again my personal recommendation): a coding language that you are comfortable with (focus on one for the beginning. Most used in the field: Python or R), solid math or statistical knowledge, a general idea about techniques to be applied in data science, and some example projects to present during an interview (in case you have no real business projects, it can be another data science project with a hypothetical link to a business context). Some great data-sets to work with can be found on Kaggle as a start, or simply use google's data-set search. But there are plenty of other publicly available data-sets, like shown in this video investigating the rise in knife crime in London.
I would strongly recommend to practice your coding skills regularly (at least two days per week). Practice by solving short tasks, such as on Codewars, DataCamp, Excercism etc.
Particularly, interesting about Excercism are their mentored language tracks, where you will get feedback from experienced programmers on your coding solutions.
If you ever get stuck, 'stackoverflow' is 'an absolute must' to search for answers, or even better, contribute yourself by posting questions or trying to help answer some. There is a high chance that a coding challenge will be part of your data science job interview. So make sure you are prepared. Have a look at some additional literature that might help , or have a look at this helpful blog  providing some questions and advice.
Refresh your stats knowledge. A friend of mine recommended the book 'Think Stats'  as a good introduction to probability and statistics using python programming. In case R will be your language of choice you could start with 'Introductory statistics with R' . But there are plenty more. Again, you might also find helpful courses on Udacity, Coursera and the like.
To boost your data science knowledge have a look at Kaggle. Not only have they lots of free data sets to play with and the possibility to enter some data science challenges, but they also have good tutorials for starters. Another free R-based course 'Statistical Learning' (main focus regression and classification) , that I can highly recommend, accompanies the free book 'An Introduction to Statistical Learning (Applications with R)' . Packt publishing offers a good range of data science books focusing on Python  and R 12
Finally, in case you are searching for some free programming books: https://github.com/EbookFoundation/free-programming-books/blob/master/free-programming-books.md#r
- Blog: the-essential-skills-and-traits-of-an-expert-data-scientist
- Blog: what-makes-great-data-scientist
- Blog: what-makes-a-good-data-scientist-at-a-small-company
- Blog: the-importance-of-soft-skills-in-data-science-book-recommendations
- Book: Cracking the Coding Interview - 6th edition. Gayle Laakmann McDowell (2015). (heavy on algorithms)
- Blog: notes-and-technical-questions-from-interviewing-as-a-data-scientist-in-2018
- Book: Think Stats - Probability and Statistics for Programmers, 2nd edition. (2014) A.B. Downey. O'Reilly
- Book: Introductory Statistics with R, 2nd edition. (2008) P. Dalgaard. Springer
- Course: Statistical Learning. (2016) T. Hastie & R. Tibshirani. Stanford University
- Book: An Introduction to Statistical Learning. (2017) G. James, D. Witten, T. Hastie & R. Tibshirani. Springer
- Book: Building machine learning systems with python, 3rd edition (2018) W. Richard et al.. Packt
- Book: Machine Learning with R, 3rd edition. (2019) B.Lantz. Packt