Steve's Blog (for Jobseekers & those hiring)

Hello,

This is an excerpt from a standard message I send candidates that are interviewing with our team for ML/AI/Data Science roles at Walmart Global Tech. I think some of the content is very specific to our company, but much of the rest could apply to roles with any company that are hiring for tech roles and leveraging “virtual” vetting through video conferencing and Zoom. Enjoy!

– Steve

Here are interview tips I have passed along to candidates for ML/AI / Data Science interviews.  Hopefully you find these useful.

Please keep in mind these are not comprehensive, but should at least provide you with a place to start to help you prepare for conversations with our team.

Tech. Phone Interview Tips (for Data Science / Machine Learning roles):

1. We leverage a code-sharing platform (most use HackerRank for Code Pairing or Coderpad or others) and connect with you on Zoom for video collaboration.  You will need access to both a computer (desktop or laptop) with strong, stable internet access, microphone/speaker enabled, and ideally a video camera, and a space with little to no noise or distractions.
2. The code-sharing platforms allows for real-time collaboration (meaning the evaluator will see your data entry in real-time, and can tell whether you are toggling back and forth between windows and whether you are copying and pasting from other elsewhere).
3. ANSWER EACH QUESTION.  (I can’t tell you how many times evaluators have told me that candidates have gone off on a tangent and talked in generalities as opposed to responding to the question posed).  After you answer, ask if that is enough detail, or do you need to expand further.  This helps you to keep from rambling.
4. Budget your time properly.  Take enough time to analyze the problem, gather needed data, answer questions directly, try not to eat up your entire time on answering just one question (keep responses succinct and to the point), elaborate further as needed, and know that the evaluator is usually trying to get through 2 or more questions in the time allotted. 

(BEST PRACTICE: Ask upfront how many questions you need to cover with each evaluator, so that you can ensure you plan your response time accordingly. – E.g: If you have 45 min. to respond to 5 questions, you know you need at least 9 min. each in a 45 min. session).

5. Re: Subject matter and content, I can’t predict the exact questions each person will ask, but they will touch on:

• CS fundamentals, including programming / coding questions, algorithmic design, applied machine learning, data structures, system design, and problem solving (much of what will be covered will relate to these areas)
• Coding – given a problem set, write clean code that will work when tested.  (write clean code that will run in production with a minimum number of bugs)  Focus on fluency / formulation.  Are you able to translate the formulations into pseudocode.  If your code does run into problems (bugs, glitches, etc.), how do you solve for these?
• Fundamentals of algorithmic development  (What is ARIMA?)
• Business Acumen / Practical Application – How to connect business problems to statistics and / or machine learning models.
• Solution / System Design – Extrapolate from provided criteria, how to solve the problem (coding, algo design, etc.)
• Not enough to get the “right” answer – you may need to outline how you solved the problem, why use a specific approach, or why not use a different approach.  Is there more than the one solution?

6. We pay attention to energy level, enthusiasm, composure, and preparation (or lack thereof) that each candidate takes.  Please prepare accordingly prior to the interviews.
7. Remember to: sit up straight, don’t slouch, don’t lean, don’t look off to the wall or elsewhere (think of the evaluator being right there in the room with you).  This is not the time to MULTI-TASK. 
8. Maintain eye contact as long as possible.  Remember: look at the camera as opposed to their image on the screen.  This gives the impression that you are talking directly to them.
9. Time permitting, ask questions (hopefully you have some prepared in advance).  Ask them about their experience with the team, advice they can share with a newbie, what they would have liked to known before joining, etc.  (This helps build rapport, allows you info on the team you may consider working with, and helps you stay “memorable” in each evaluator’s mind).
10. For further prep., I know our Sourcers have recommended these links:

                                                    i.     HackerRank: https://www.hackerrank.com/interview/interview-preparation-kit

                                                   ii.     LeetCode: https://www.leetcode.com

                                                  iii.     Secrets to a Successful Interview https://medium.com/walmartlabs/secrets-to-a-successful-data-science-interview-73e317a4d620

NOTE: THESE MAY OR MAY NOT APPLY TO VERY SENIOR LEVEL ROLES: 

11. Listen for hints, clues and other help that the evaluator may provide.  Sometimes, if you are close, they may be willing to provide that nudge to get you going in the right direction.
12. Even if you do not provide the right answer, but can provide justification and an indication that you put some thought into your response, our team may look on that favorably.