Landing you data science job – Part 2: Getting ready for your interview
Want to learn how to prepare for your data science application? Read our first part here.
General preparation for interviews:
– The 120 Data Science Interview Questions. If you want the physical questions of the book, you can get it here for $19, but the answer themselves are free on Quora Learn Data Science
– Springboard 109 data science interview questions, it is divided into six categories: Statistics, Programming, Modelling, Behavioral, Culture Fit and Problem Solving
– The Data Science Interview: Personalised interview questions for the positions you want to apply for, and you can see also what others people are answering to those questions or view popular questions asked at companies like Facebook, Google, Amazon.
– Machine Learning Cheatsheet: Brief visualisations of different machine learning concepts with diagrams and code examples.
4. The Live Coding Interview
Your CVs have impressed your employers (or the CV Screening algorithm), now it is time to prepare for the online coding challenges.
Coding challenges can range from a simple Fizzbuzz question to more complicated problems like building a time series forecasting model using messy data. These challenges will be timed (ranging anywhere from 30mins to one week) based on how complicated the questions are. Challenges are usually hosted on sites such as HackerRank, CoderByte, or even internal company solutions.
More often than not, you’ll be provided with written test cases that will tell you if you’ve passed or failed a question. This will typically consider both correctness as well as complexity (i.e. how long did it take to run your code). If you’re not provided with tests, it’s a good idea to write your own. With data science coding challenges you may even encounter multiple-choice questions on statistics, so make sure you ask your recruiter what exactly you’ll be tested on.
Keep in mind that companies not only looking for your ability to solve the problem, they also look for code readability, well-designed and clear functions and sometimes the optimal solution for a specific challenge. After finishing your code and pass the challenge, you should go back, improve the readability and optimise the code.
Preparation resources:
- For SQL and data structures/algorithm questions: Leetcode
- Review for math and statistics questions Brilliant
- More SQL exercises: SQL Zoo and Mode Analytics
Tips:
- Skim read all the questions and rank them from easiest to hardest. This will allow you to estimate the time taken to do each question while your subconscious mind started working on many questions at the same time and scheduling the questions in order.
- Start with the hardest problem first to fully understand some them and when you hit the wall, move to the more straightforward questions before returning to the harder one.
- Pass all the test cases first, and after you finished, go back to improve complexity and readability.
- If you’re done and have a few minutes left, get up, walk around a bit and get a drink before getting back to your desk. Read your answers once last time and then submit it. You might find some small tweaks in your code after that cup of tea.
Things to avoid:
- Don’t dwell on questions. Set the time limit for each question and if you hit the time, move on.
- Write the whole code with no readability ( a.k.a. long code, no comment). Even though you have solved the problems, adding some comments will certainly a great way to communicate your ideas to interviewers and help them to understand your logic.
- Don’t prepare: Prepare for the interview questions with Cracking the coding interview (best coding preparation of all time), HackerRank or Leetcode. It will be amazingly beneficial for your next-round white-board or in-house interview as well. ( or even software engineer interview)
5. Telephone screening interview
Your CV has successfully impressed the hiring managers or the HR recruiting team, and now you will have an exciting first interaction with your future employers. Depending on the company, there can be one person from HR or your hiring manager, or both. You should always ask for whoever is that may interview you and even find your interviewer’s Linkedin. If it is only the HR manager, they will only ask about general questions about your resume, why you want to apply for this role and examples of why you think you may be a good fit for this role based on your experience and personal character. With a hiring manager, you should expect a few more technical questions such as SQL or some algorithm questions. In the end, you will be given a few minutes to ask relevant questions.
Tips:
- Do your company homework
- Reach out and ask who is going to interview you and what the interview is about ( do you have to prepare for technical questions?)
- Read your CV again before the interview: Make sure you will come as precise and clear about
- Prepare for relevant questions: What is your motivation?
Things to avoid:
- Don’t ask questions. HR aims to check for your background, motivation and culture fit. Asking questions that can’t be found in the public domain about the role, the culture, the team and the company will show genuine interest in the company.
- Ask irrelevant questions
- Don’t speak anything negative about any of your past employers.
6. The take-home project.
Take-home projects are becoming increasingly popular, as it resembles the real-life projects data scientists are going to work on. They usually happen between the Telephone Interview and the on-site interview, and you can expect to present your findings on the on-site interviews. Questions can range from general( here is your data, tell me some insights you can get out of it) to specific( build me a clustering model out of this data set). You are also expected to create a presentation to present your findings to hiring managers or business stakeholders.
Tips:
– Ask as many questions as possible to clarify what you are being tested on and what they are expected to listen.
– Email to ask if you need any clarification, and if not, you can make assumptions, but remember to state them clearly.
-Ask to know who are your audience. If you are expected to present it to business executives, avoid technical jargons and focus on actionable insights, while you can put that in if your audience is more technical.
-Practice this take-home challenges: you can view this super helpful answers(without dataset) from JifuZhao GitHub repo.
Things to avoid:
-Ask no questions: Companies normally give general, ambiguous task( e.g. give me insights out of this dataset) without giving any context. If you don’t ask any questions, your insights when you present it may not be what they expect, so please, ask questions.
-Take too much time to do the project: A take-home project can take anywhere from 3 hours(according to employees, it normally means at least 6 hours) to a few days and most of the time, if you fail, they will give limited( general) feedback. Try your best but don’t dwell on one project, if the project takes too long, focus on other applications.
-Don’t keep the project: you are very likely to be asked about the project on the on-site interview, or when you started the job( the project is directly related to the job most of the time). So keep the project handy
7.In-person interviews(on-site interview)
This process normally contains a series of interviews throughout the day and you will be likely to meet many people from team members to hiring managers and sometimes CTO depending on the size of the company. It is worthwhile to remember at this stage; the company want you to succeed. They have spent significant resources and time to source your application to interview a handful of candidates to choose only a small number to visit their companies, so be confident and be prepared!
Tips:
-Ask who is going to interview you and research them: It is great to know who is going to interview you to prepare for potential questions they may ask. A PhD senior data scientist will ask different questions than a data analyst or company executives. As mentioned above, the questions will depend on who will interview you, so have an idea to what type of questions are they going to ask will be helpful.
-Do you research about the company: from culture to interview questions: Glassdoor, Crunchbase can be very helpful
-Prepare common interview questions: Behavioral Questions such as “Tell me about yourself” and “What is your plan in 5 years?” are common initiating questions. Prepare short, concise answers to show interviewers that you are well-prepared.
Things to avoid:
-Dress too casually: Even though data scientist is a highly technical role, but the first impression lasts. It is always wise to dress one level above the company’s dress code (e.g. if everyone dress casual in jeans and a T-shirt, you should go for business casual)
-Don’t prepare your company research and questions: At the end of the interviews, you will have a chance to ask questions about the company. If you don’t have any questions, this will show a lack of interest in the company.
8.Post Interview Steps
Congratulations! You have now nearly finished with the interview process. After the interview, it is always the best practice to send interviewers thank-you notes on 12-1 pm the day after that. This will show you are genuinely interested in the position at the company, and it will make interviewers recall and remember you better.
If you have many job offers, you should check this job offer negotiation article by Dataquest.
Good luck with your job hunting!
For more data science opportunities, view our current opportunities on our website or contact our senior account manager, Dan. Follow our Blogs, Twitter or Linkedin for the latest updates on the data scientist news, the disruption brought by the adoption of AI and some recruiting tips.
