Source: towardsdatascience.com
I have summarised in this text best tips for data scientists to progress along their career. Whether you’re just starting or you want to go from junior to mid or mid to senior, there’s something for you.
Portfolio of GitHub projects
First of all you should build a portfolio of open-source projects on GitHub. I recommend to create three projects:
- A classification project where you use a public database (you can download one from Kaggle — more about that in the next paragraph) of images/texts to sort them and hone your skills with supervised/unsupersived learning (from PCA to neural networks, through DBScan, KNN, etc.).
- An NLP project which would analyse sentiments from Tweets from a particular subject and classify them accordingly into positive/neutral/negative. That’s a classic one — choose a topic interesting to you, so you have a good story to tell about it.
- A scraping project, where you scrape different sources to extract information — that can be scraping sports news if you’re a sports fans, financial fluctuaction if you’re into finance or data science novelties. The endgoal might be creating an automated website with scraped and extracted content. Easy to show during a job interview.
If you’re more advanced then you definitely should play around with latest machine learning algorithms. For example you can:
- try GANs (Generative Adversial Networks) and generated some faces or cats.
- try Reinforcement Learning with easier games.
- try GPT-2 and text generation.
You have endless opportunities here. You definitely should show that you’re well-versed in neural networks and their fundamentals like Keras, PyTorch, TensorFlow.
Know what you know and what you don’t know
Even if that sounds obvious you should be able to answer questions about your knowledge, particular technicalities or problems.
Actually the most important question you’ll be asked is what was the hardest in a given project. If you were struggling with a particular technical issue, why and what it was. How did you overcome it? Those are the basic kind of questions that you should be prepared to answer.
Be ready to talk about algorithms in more detail, different methods you have used in the past. Be open and share also what was problematic for you. This can only help during a job interview.
Polish your LinkedIn profile
The last thing which is often overlooked by data scientists is putting up a coherent LinkedIn profile with explanation of what you did in the past and where you’re at currently. If you have any gaps in your career be sure to be asked about them — there’s nothing wrong with 6-months stay on Bali surfing, you should just be honest about what motivated you back then and what you wanted to achieve. I imagine that a good argument would be you wanted to have a break from your screen, wanted a restart, wanting to work as a digital nomads — there are plenty of reasons to explain why this option was the best also for your career.
You should also ask for recommendations from your past employers. They can write it directly on LinkedIn — just a couple of sentences is enough. If your last departure wasn’t really planned and you feel bitter about, try to explain why and what you expect from your next employer.
You should be able to say precisely why you want to change your job. Is it because you’re looking for new challenges? If yes, why you can’t have them at your current job? Speaking clearly about your previous work experiences is a huge asset.
Shine!
Summing up there are 3 things you should do to really master a job interview for a data scientist position:
- build a porfolio of GitHub projects;
- know what was the hardest part of each of your project and how you overcame it;
- polish your LinkedIn profile.
Good luck!
And if you want to know more, read my other articles about becoming a Data Scientist:
- 5 ways to become a data scientist
- How to start with Data Science
- 3 common mistakes Data Scientists make
- Practical guide to become a Data Scientist
