SDS 615: How to Ace Your Data Science Interview

Podcast Guest: Nick Singh

October 4, 2022

Data scientist: Nick Singh wants to help you sell yourself. In this episode, Jon Krohn and his guest discuss how to ace the data science interview, methods to build a job-winning portfolio, and why you need to stop thinking that talking about your work is “shameless promotion”.

Any questions? You should have them, and Nick will explain why later on in the show!

Thanks to our Sponsors: 
Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information.
About Nick Singh
Nick Singh is an Ex-Facebook & Google Engineer turned best-selling author of Ace the Data Science Interview, and founder of SQL Interview Platform DataLemur.com. His career advice on LinkedIn has earned him 100,000 followers, and he’s successfully career coached 578 people to land their dream job in data!
Overview
You have an education in data science. You have analytics experience. Peers promised you that companies would be salivating at the very thought of your data mining and visualization skills. Why is it so hard, then, to get past the interview stage? This week’s guest Nick Singh breaks down the common roadblocks to acing the interview and how he found a surprising gap in the market for tools that help data scientists through the recruitment process.
Nick explains how being flexible in what he had to offer the business of data science helped him become an author, educator and business founder (for more on how to develop this type of bravery for your own career, we recommend listening to episode 610: Who Dares Wins!) With a LinkedIn following that numbers in the hundreds of thousands, Nick is no stranger to fan feedback, and he listens: Readers of his bestselling book Ace the Data Science Interview wanted to be able to put what they learned into practice. Their feedback inspired him to found DataLemur, a free SQL interview platform on which users can quiz themselves with industry- and even brand-specific questions. DataLemur strengthens users’ knowledge with questions that are ranked by difficulty. Registered users can write their own SQL queries on the platform, for which they get personal feedback from their peers.
By responding to a simple need – to help data scientists get the job they want – Nick has established a brand with multiple applications. In addition to the book and platform, he has also launched a course called “Ace the Data Job Hunt”, which focuses on the non-technical aspects of the job search that typically get forgotten. From resume formatting to behavioral interviews, Nick walks through the methods to finding and landing a coveted position at a MAMAA (big 5 tech) company, using real-life examples from his interviews at Facebook.
And it’s not all work-and-no-play for Nick Singh! During the show, Nick also talks about moonlighting as a DJ, and how channeling your creative energy can help you to get ahead in the business world.  
In this episode you will learn:   
  • Nick’s inspiration for writing his bestselling book, Ace the Data Science Interview [06:21]
  • Why Nick believes in being a work generalist [12:37]
  • How DataLemur supports emerging data scientists for free [15:43]
  • Why Nick started DataLemur off the back of his book [21:31]
  • Portfolio essentials for any data scientist [22:36]
  • The three most common things data scientists get wrong at the interview [24:33]
  • How data science introverts can shift their mindset about self-promotion [37:58]
  • Great responses to end your data science interview on the right foot [42:21]
Items mentioned in this podcast:
Follow Nick:

Podcast Transcript

Jon Krohn: 00:00

This is episode number 615 with Nick Singh, bestselling author of the book Ace the Data Science Interview. Today’s episode is brought to you by Datalore, the collaborative data science platform, and by Zencastr, the easiest way to make high quality podcasts. 
00:19
Welcome to the Super Data Science Podcast, the most listened to podcast in the data science industry. Each week, we bring you inspiring people and ideas to help you build a successful career in data science. I’m your host, Jon Krohn. Thanks for joining me today. And now, let’s make the complex simple. 
00:50 Welcome back to the Super Data Science Podcast. We’ve got a special episode for those of you out there who are looking to land a data science job for the first time, level up into a more senior data science role, or perhaps land a data science gig at a new firm. Our guru guest for all your success in data science interviews is Nick Singh. Nick co-authored the bestselling book, Ace the Data Science Interview, an interview question guide that has sold over 16,000 copies since it was released last year. He also created the DataLemur platform for interactively practicing interview questions involving SQL queries. He previously worked as a software engineer at Facebook, Google and Microsoft. He holds a bachelor’s in engineering from the University of Virginia.
01:34 In this episode, Nick details his top tips for success in data science interviews, common misconceptions about data science interviews, how to become comfortable with self-promotion and increase your chances of landing your dream job, strategies for when interviewers ask you if you have questions for them, and the subject areas and skills you should master before heading into a data science interview. All right, you ready for this highly practical episode? Let’s go.
02:06 Nick Singh, welcome to the Super Data Science Podcast. I’m delighted to have you here. Welcome. Where are you calling in from? 
Nick Singh: 02:13
Thank you for having me, man. I’m in Washington, DC. I’m in the suburbs in Northern Virginia. 
Jon Krohn: 02:18
Oh, nice. So in New York today, just, what is it? Like a two hour ride on the Acela train away from you. 
Nick Singh: 02:28
Something like that. 
Jon Krohn: 02:28
I do love the Acela train for getting between New York and Washington, DC. For people who don’t live in North America that are listeners, you probably live in a place that has great public transport. And in North America, we just have so little of it, but one of the few exceptions, the only high speed rail line in North America connects Boston to New York to Philadelphia, to Washington, DC. So it’s this straight north, south line along the east coast of the US. And it’s actually really nice.
Nick Singh: 03:03
And it works. Yeah, it works. And there’s no security or very minimal so you can just kind of hop in, hop off. 
Jon Krohn: 03:09
Yeah, exactly. So it’s way less stressful than a flight. They serve you great food. You can walk around. I love it. And the reason why I mentioned that is that today in New York it was a beautiful, hot day, we’re recording in mid-September. And so I imagine you had just a wonderful, brilliant day in Washington, DC today. 
Nick Singh: 03:31
Yeah, no. I went out, I went on a little run. So we’re having a good day and it’s a great way to end a Friday evening. 
Jon Krohn: 03:37
Me too, man. It was like, ah, a day like today, we may not have that many more of these this year, so I went out for a run before this too. Maybe that’s why I’m feeling so good and why you’ve been… Listener, I think you’re in for a great episode here because Nick and I have been talking for an hour before we started rolling. Him and I just have so much to say and we’re in such a good mood. So the weather has really blessed this episode. So we know each other through LinkedIn, basically. So Nick, in February, I don’t know who added who on LinkedIn, but then Nick messaged me back in February of this year, asking if he could be on the show. 
 
04:21
And I got back to him. I loved your profile. I thought, a guy who’s written a book on getting a data science interview, that’s awesome. I’m sure we have lots of listeners out there that would love to hear about an expert on data science interviews. And so I asked Nick if he could send me examples of talks that he’d done, like previous podcasts, appearances. It turns out Nick has done tons of them, including in mainstream media, but he didn’t reply to my message. 
Nick Singh: 04:49
Things fall through the cracks. So I’m ambitious and then things happen, and then, you know how it goes. 
Jon Krohn: 04:55
I do, but it doesn’t matter because now you’re here. And the reason specifically how you ended up here is some regular listeners will know that we have a brilliant researcher on the show. His name is Serg Masis and he’s actually even been a guest on the show. So he was a guest on episode number 539, talking about interpretable machine learning, which he is an expert in, which he wrote an outstanding book on. And so super lucky to have Serg working for us on the team. And a couple of weeks ago I had this idea. I was like, wow, Serg is super well-connected in the data science world. He’s constantly on social media, has a great impression of who the great content creators are in data science. 
 
05:38
So I finally have this brainwave after many months of Serg working for us. I said, “Hey Serg, if you ever have any ideas for guests that should be on the show, let me know.” And he wrote back immediately. He was like, “Ah, I’m so glad you asked because I’ve made a list. I’d already had this list prepared.” And so over the next few months, we’re probably going to have lots of guests that were ideas from Serg. And then there were a couple of people on this list that he put an asterisk next to as somebody that you should try to get on the air as soon as possible because they’re really brilliant. And Nick Singh, you were one of those names. 
Nick Singh: 06:14
I appreciate it. And I appreciate you, Serg. So thanks for the shout out. 
Jon Krohn: 06:19
So Nick, you wrote your bestselling book, Ace the Data Science Interview during the height of the pandemic. Could you walk me through how the idea for writing this book came about? Why did you think it was a good idea? 
Nick Singh: 06:32
So me and my co-author Kevin Huo, who we worked together at Facebook and we were actually long-time friends, we knew that there was space in helping data scientists prepare for the interview and land jobs, because in the software engineering world, which Kevin and I are familiar with, because we both also studied computer science in undergrad, there’s books like Cracking the Coding Interview and there’s platforms like LeetCode. But there was this big gap in helping data analysts and data scientists and machine learning engineers with their interviews. So we knew there must be something there and there wasn’t. 
Jon Krohn: 07:09
It’s such a great opportunity. So for software engineers, they have Cracking the Coding Interview. And so this seems to me, it’s kind of like the analogous guide for data science. 
Nick Singh: 07:18
That’s what we were hoping for. And one year in, with 16,000 copies sold and it being a number one seller in several Amazon categories. So yeah, the equivalent wasn’t there for data science. And we were like, well, someone should do something. And then after a while we were like, maybe we should do something. We’ve been able to work at some of these top Silicon Valley companies and know a thing or two about data. Why don’t we throw our hand at it? And I think COVID really helped us. I mean, it was a terrible thing for the world and it affected our lives too, but one thing it really did was put us inside in our home. So now Kevin and I couldn’t go out, we couldn’t hang out with our friends and we are long-time friends. 
 
07:58
So we couldn’t hang out with our friends, we couldn’t do stuff. So that was the right point to go write this book. And then something else that’s really funny, which is I moved out of San Francisco and came back to live with my parents, because that was a lot of what people did in their mid-twenties when, hey, why pay this expensive rent if you can’t go out and do anything? And I had this realization. I loved my job where I worked last. It was great. But there was just some realization that, hey, I’m earning money. I’m not paying rent. I’m not really paying for food, I’m eating my mom’s food. I can’t do anything. Where’s my money going? And then secondly, I’ve always wanted to do my own thing, whether it’s a book or start up, something of my own. And I was like, hey, look, if I don’t do it now, when I don’t have rent to make, I’m not getting any younger. 
 
08:48
I don’t have kids. I don’t have a mortgage. If I don’t do it now, I will probably never do it and I’ll always regret and say, this was the right time in my life. I could have done something and I didn’t do it. So that was the big thing that pushed me there, which was just seeing future Nick at age 30 or 40 or 50 being like, “Hey, you had that chance at 24, 25 when the world kind of shut down and you didn’t take it.” So that’s kind of what really pushed Kevin and I to get the book done. Because we had that idea for a while, but you know how book ideas are. We’re always having them but no one’s ever doing it. But COVID was the time we got it done. 
Jon Krohn: 09:21
It’s a huge amount of work. And there’s two things that you said that really resonate with me, that I have analogous experiences in my life. So similarly, the COVID pandemic hitting gave me a great opportunity to create a huge body of content on the mathematical foundations of machine learning. So linear algebra, calculus, probability theory, statistics, algorithms and data structures. What are the key things that you need to know in those subject areas in order to be a great machine learning practitioner, in order to be a great data scientist? And similar to you, I’d had this idea for a long time of creating that content. 
 
09:55
And then it was the pandemic hitting that I was like, well, not only is there a whole bunch more time because I can’t go out and have fun, I can’t go on trips, but in addition, for me, it was also a coping mechanism. So digging deep into a linear algebra textbook was a great escape from a hugely disorienting time for me, where at the beginning of the pandemic, we didn’t know it was going to be only a little more than a year before we were going to have the first vaccines rolled out. It was potentially years of being stuck indoors. 
Nick Singh: 10:32
And in New York, you were in New York so it must have felt even more… Because it’s a big city and you guys got hit really hard in the beginning. 
Jon Krohn: 10:40
Yeah, it was super crazy here for sure. And the other thing that you said that really resonated with me was taking the opportunity that you have now, where you don’t have dependents, to do your own thing. I had the same experience myself. In 2015, I had the opportunity to join an early stage startup as their chief data scientist. I left a very comfortable corporate job where things were going super well and I had no real intention of leaving, but I can corroborate what you’re saying. And so if there’s a listener out there that you’re thinking… A lot of the barriers to you being successful as an entrepreneur or having a successful side hustle, pretty much all of the barriers in most cases are just mental. 
 
11:36
And so in general, regardless of what your life circumstances are, I encourage you to think about if you have a dream of having your own thing, figure out a way that you can be doing it for an hour or two a week to start, just get [inaudible 00:11:51], block it off on your calendar and do that. But specifically, that experience you’re describing, so I guess I was about 30, maybe 29 when that opportunity happened and it was kind of scary to me. I was like, wow, I’m leaving a great career trajectory in this comfortable large corporation to be going off with a company that only has a certain amount of runway, and then who knows what happens after that? But same thing as you, I was like, if I don’t do this now, when I have no dependents, when is going to be a better time? So I guess especially if you’re in that situation, listener, you might want to jump on it while you can. 
Nick Singh: 12:34
Agreed. Agreed. 
Jon Krohn: 12:36
So super cool. So now you define yourself as a career coach, right, Nick? So previously, as you’ve alluded to, you were a software engineer at several big tech companies, basically all of the biggest ones, Microsoft, Google, Facebook. And so you were a software engineer and now you define yourself as a career coach. 
Nick Singh: 13:01
Yeah. 
Jon Krohn: 13:01
So how did you realize that this, being a career coach, is what you were meant to be doing? 
Nick Singh: 13:08
And actually it’s really funny because you see software engineer, I see, well, one summer I was back end, one summer I did full stack. I did growth and then I also did data engineering. I did a data science internship. In school, I studied computer science and systems engineering. And systems engineering is basically a mix for… In other schools it might be called industrial engineering or operations research, but it’s basically a mix of some math, some statistics, some business, some just general engineering skills applied to hard engineering problems in factories and queuing theory. It’s just like, that’s a vague thing. My life has been a lot about being a generalist. And that’s why, when you say that I’m a career coach, I kind of laugh because, well, you’ll hear about it. Am I an author? Am I a career coach? Am I a LinkedIn influencer? 
 
13:57
Am I a data scientist? Am I a software engineer? Am I an entrepreneur running DataLemur? Because those, right now I feel like a product manager adding features and making sure DataLemur, which we’ll talk about, is really successful. So honestly, I don’t even know if I really call myself a career coach, but just to make things easier, you can box me in whichever way you want. So I like to pick my title because some days I genuinely am an author and some days I’m like, yo, I don’t even love to write that much. I’m more of a tech product manager, running DataLemur, entrepreneur type. So I just pick what I want depending on how I feel in the day. 
Jon Krohn: 14:37
Today’s show is brought to you by Datalore, the collaborative data science platform by JetBrains. Datalore brings together three big pieces of functionality. First, it offers data science with a first-class Jupyter-notebook coding experience in all the key data science languages, Python, SQL, R, and Scala. Second, Datalore provides modern business intelligence with interactive data apps and easy ways to share your BI insights with stakeholders. And third, Datalore facilitates team productivity with live collaboration on notebooks and powerful no-code automations. To boot, with Datalore, you can do all this online in your private cloud or even on-prem. Register at datalore.online/SDS, and use the code SUPERDS for a free month of Datalore Pro, and the code SUPERDS5 for a 5% discount on the Datalore Enterprise plan. 
15:28 Nice, Nick. Well, it sounds really exciting. I love that you get to wear all those different hats in your week and that you have that kind of independence to be exploring and focusing on those different kinds of roles. Product management, career coaching, LinkedIn influencer, it all sounds really exciting. So you mentioned DataLemur there, your platform. So you have this SQL interview platform called DataLemur. I’ve used it. It is super cool. So you can log in for free, and in fact, today, a hundred percent of the functionality, correct me if I’m wrong, a hundred percent of the functionality in the platform is free. So you can create a free login. At the time of recording, the platform has only been live for two weeks, but you’ve had tremendous engagement already, thousands of users. And I can see why, because it is a super intuitive platform. So you can come in and you can select from different kinds of SQL problems. They’re ranked as easy, medium or hard. A lot of them are related to chapters of your Ace the Data Science Interview book, and they’re specific to questions at specific companies. 
 
16:39
So big tech companies or really cool startups like Robinhood. And so it has real data in the back end, real data tables, and then you write your SQL queries to extract data from those tables and there’s a code executing terminal. And in that terminal, you can write real SQL code and then you get the output of the results. And so there are platforms that do that kind of thing, there are other educational platforms that do that kind of thing, but this is the first that I’m aware of that is doing it specifically for SQL interview questions. For a platform that is so new, you’ve done an excellent job, Nick, of having it work intuitively and efficiently. 
 
17:26
It’s a joy to use. And then one of the other features that I love about it is that if you don’t want to skip… Let’s say you’re struggling to figure out how to solve a problem, but you don’t want to jump right to the answer because then you’ve spoiled the opportunity to learn, to some extent. You’ve offered several tiers of hints. So it sounds like on many questions you have up to four tiers of hints that you can ask for. 
Nick Singh: 17:55
Every question has at least one hint and most have three, maybe four hints. And it’s because there are other SQL interview platforms like LeetCode and HackerRank. And they’re pretty good, but in my opinion, a little clunky to use. And secondly, they’re kind of competitive. There’s this element of getting points and a leaderboard. And that’s great, but for learning, I want to give people easy access to solutions, I want to give them easy access to hints and I want to make it all free. So that’s basically kind of the ethos of DataLemur. And it has today, 60 FAANG SQL interview questions that you can go practice. 
Jon Krohn: 18:33
Oh, yeah. So FAANG, well, buzz- 
Nick Singh: 18:36
Little buzzword there. 
Jon Krohn: 18:37
It was the buzzword certainly a few years ago to describe the prominent big tech companies. So Facebook, Apple, Amazon, Netflix, and Google, the companies at the time. It doesn’t sound as good, but I think the most current acronym is MAAMA. 
Nick Singh: 18:57
I’ve heard MANGA. It’s a free for all, so until someone really…
Jon Krohn: 19:02
Because that MANGA, was is that N, Netflix? 
Nick Singh: 19:07
Yeah. 
Jon Krohn: 19:08
So with MAAMA, it’s M-A-A-M-A. You’ve got Microsoft, Apple, Amazon, Meta, and then Alphabet, Google’s parent company. And I think those are the big five tech companies today, really. But it hasn’t caught on. FAANG was like, that was the first big one where people started doing this grouping all the big tech companies into one acronym and it worked really well. And so it still makes sense that people say that today, even though when you say FAANG today, you probably aren’t talking about Netflix. 
Nick Singh: 19:40
And we keep leaving out Microsoft, which is wrong. 
Jon Krohn: 19:45
Right, right. It’s crazy that FAANG [inaudible 00:19:47] Yeah, they’re huge. They’re sometimes the biggest company in the world in the last few years. And they also, for data science, super cutting edge. In the preceding episode of this program, we had Dr. Emre Kiciman, who’s been working at Microsoft research for 17 years. And Microsoft research is tremendous. I mean, I talked about it in that show and I haven’t looked up since to make sure that this is correct, but I believe that Microsoft publishes more papers than any other tech company, which means probably more than any other commercial organization.
Nick Singh: 20:26
I can believe that. Yeah. 
Jon Krohn: 20:30
So we shouldn’t be leaving Microsoft out. Anyway, how did we get onto this? 
Nick Singh: 20:36
Sorry. 60 interview questions on [inaudible 00:20:41]. 
Jon Krohn: 20:41
That’s right [inaudible 00:20:41]. So I love that. And then there’s also, you have a vibrant discussion section. So for any one of those 60 questions, at the time of recording, you can go and see a discussion where… I’m looking at one of the questions now, there’s hundreds. 
Nick Singh: 20:57
There might be hundreds of people discussing their solution, getting help, criticizing. Well, maybe criticizing is too heavy a word, critiquing other people’s style. And that’s a phenomenal way to learn because it’s like a free code review. 
Jon Krohn: 21:12
Yeah. So super cool, Nick. I am deeply impressed by DataLemur. And again, it is free. We’ll of course have a link to it in the show notes. So how did it occur to you? So you set off on your own, you write this bestselling book, Ace the Data Science Interview. How did it then occur to you that the next step was to create this free platform, DataLemur? 
Nick Singh: 21:42
Absolutely. So one of the biggest feedbacks we got was, “Hey, your SQL questions are awesome from the book, but I want to run this code. And you’ve just given me the table structure, but not any real values to work with. And I don’t want to mess around with my own local Workbench or DB Fiddle or this or that. Can you load up these questions in a platform so I can actually run my own query?” So it was pretty logical because already so many people had liked the book, and that was the one big repeated pain point of the book, that they said the SQL was good, but they wanted to run it. So we did it. 
Jon Krohn: 22:21
Awesome. Well, congrats on your early success of the platform. I can’t wait to see where it goes. And in addition to the book, Ace the Data Science Interview and the free SQL interview platform, DataLemur, you also have a course called Ace the Data Job Hunt. Do you want to tell us about that? 
Nick Singh: 22:40
Yeah, sure. It’s 25 video modules that elaborate on the non-technical parts of the job hunt. So the beginning parts, which are actually the first four chapters in my book, which are about resumes, portfolio projects, cold emails and behavioral interviews. So it might be funny to think, why are we focusing about this? I think that on your own, working on these data science interview questions from the book or solving the SQL questions on DataLemur, they’re really good. But there’s this really big piece about how do you get the interview in the first place? How do you set up your resume, your portfolio project, that the book addresses, but there’s just something different when I’m on video showing you how I’m roasting resumes or how I would build this portfolio project, or here’s how I send a cold email and here’s what to write. It just wasn’t translating from the book. 
 
23:32
And interestingly enough, a lot of the people said that the most interesting thing about the book were those first four chapters. Because, sure, I like my book, I think it’s pretty good. But there’s a ton of other great machine learning books. So the machine learning interview questions are good, the solutions are good, but it’s not like I’m the first person to break down machine learning. Yet for a lot of these concepts, when I’m actually showing you the cold emails I sent to get interviews at Airbnb and Snapchat, or the portfolio project that I used that helped me get my job at Facebook, or the exact resume Kevin and I actually used when we got to Facebook, and what, in retrospect, we could have improved and what we think it did well, that kind of thing, you can’t translate in a book that easily, but on video it really gets the point across. 
Jon Krohn: 24:21
Very cool. Love it, Nick. So there’s another resource for you to check out, listener, Ace the Data Job Hunt, that’s Nick’s course. So there’s a lot of career advice online regarding the hiring process, particularly with the tech giants, the FAANG or MAAMA companies, do you have guidance for listeners on what some of the most common misunderstandings are in this kind of career advice? 
Nick Singh: 24:53
So when it comes to resumes, a big misunderstanding is that more is better, when you don’t realize that these recruiters and these hiring managers are so inundated with resumes, actually having a simple plain text resume is better than these colorful well-designed resumes that are 2, 3, 4 pages, that list out all kinds of crap. So that’s the first big thing that people don’t realize. The second one is around certifications. They’re really not a big deal at the top tiers and the top tech companies, because, honestly, these certifications aren’t that challenging to get, because the companies that are running the certification, let’s say some Snowflake certification or AWS certification, it’s sort of like their marketing and branding thing. 
 
25:40
So they kind of want more people to be AWS certified so they don’t have an incentive to make the test hard. So conversely, at these top tech companies, they don’t even care about them because they’re not really that real or that important. So I’ve seen a lot of people, in particular, the Google data analyst certificate that they have out there, think that, oh, because it has Google on the name they’re going to get shortlisted at Google or a Google comparable company. It doesn’t really work that way. 
Jon Krohn: 26:08
Yeah. Couldn’t agree more. Agree a hundred percent, Nick. 
Nick Singh: 26:10
The third big misconception people have is that interviewing is the same as being a great data scientist, and actually they’re two different skills. And people get this kind of complex that, hey, you know what? I have eight years experience. I have 10 years experience, I have a PhD. I don’t work with SQL, I do Pandas all day. Do I really need to prepare for the SQL interview? And here’s the sad truth, which is, hey, these companies run really rigid processes that have thousands of people go through them. So they can’t really tailor it that much and say, “Oh, you know what? You use Pandas? We’ll ignore SQL,” or, “Oh, you have eight years experience? You’re above this question.” 
 
26:50
They don’t really do that. So sometimes that means a new grad might have more experience with a question around statistics 101 because they took that class recently and you might be a little hazy. Some people have this kind of complex, like, “Oh, if I have to study for the interview, it’s not a job I want,” or “Hey, if I have to go out of my way to do SQL or review some probability and statistics that I don’t necessarily do in my everyday day-to-day, that I should still be entitled to these high-paying jobs.” And the truth is, if you don’t want them, that’s fine, but if you do want them, you got to play the game. And that’s what the game is. It’s working through the questions that they ask and not trying to make up your own reasoning for why you’re above that. 
Jon Krohn: 27:34
Perfectly said. Another example of this happens in investment banking. So if you want to work at one of the top Wall Street firms as an investment banker, those investment bankers in their downtime, so sometimes they spend… They might be at work for hours with nothing to do in the middle of the morning or the middle of the afternoon. And then something can drop on their desk at 6:00 PM and they have to be there until 6:00 AM working on some pitch deck for some proposed merger or acquisition or something. So in that downtime, while they’re sitting around, something that they do is they go through… There’s kind of similar to the Ace the Data Science Interview book that you’ve written, Nick, there’s Wall Street guides like that for Wall Street jobs. And so they’ll just ask each other, they’ll just work through those books to keep each other sharp on the questions so that they can be interviewing well at other firms and just keep their options open. 
Nick Singh: 28:28
No, that’s savage, practicing for interviews at your own employer with your coworkers. I love that. 
Jon Krohn: 28:37
So, cool. Those were great misconceptions to bring up for our listeners. So the first is that more is not better on your resume. The second is that certifications are not a big deal. And the third is that being a great data scientist does not mean that you’re necessarily a great data science interviewee. You’ve got to keep sharp on the data science interview skills with books like yours, Nick. So there you go. So it’s a resource, not just for people early in their careers, but also people who are established who might be thinking about moving. Awesome. So do you think that some of these issues, maybe even issues around these misconceptions, so issues around big tech interviews, do you think that big tech companies will catch on to that and maybe in the future adapt their interview process so that they say, “Okay, look, this person does have eight years of experience? It’s crazy that we’re quizzing them on the fundamentals of SQL. We should be testing them on leadership skills or their ability to get papers published or something like that.”
 
Nick Singh: 29:44
A little bit, but I don’t really have that much hope for the change. The one part I do see, that I’ve actively seen, is that some companies are saying, “Hey, look, I have this data manipulation problem. Whether you want to use Pandas or dplR and R, or SQL, go for it.” But there’s still enough companies that are pure on SQL only. And they say, “Hey, tough luck. This is SQL screen. This is what the boss’s boss’s boss said was part of the round, whatever.” So I don’t have too much hope for things changing. And I think that there are reasons. Having worked at them I’ve seen how many people are applying to these jobs and what it takes. There are definitely reasons why they have orchestrated it the way they have. So I’m not really too sure about it changing. 
Jon Krohn: 30:32
Trying to create studio quality podcast episodes remotely used to be a big challenge for us, with lots of separate applications involved. So when I took over as host of Super Data Science, I immediately switched us to recording with Zencastr. Zencastr not only dramatically simplified the recording process, we now just use one simple web app, it also dramatically increased the quality of our recordings. Zencastr records lossless audio in up to 4K video and then asynchronously uploads these flawless media files to the cloud. This means that internet hiccups have zero impact on the finished product that you enjoy. To have recordings as high quality as Super Data Science yourself, go to Zencastr.com/pricing and use the code SDS to get 30% off your first three months of Zencastr Professional. It’s time for you to share your story. 
 
31:21
I think one of the reasons why it’s hard to get rid of these quantifiably measurable tests is that they’re unbiased, or they minimize bias at least. So if you were to say, oh, for people who have eight years of experience, we won’t give them the SQL test or any of these other tests where there’s an objective right answer or wrong answer. We’ll just kind of get a feel for how they interview and what we like about them. That won’t fly with the HR teams at these big tech companies, because it’s really important to those HR teams to make sure that people’s implicit, or worse, explicit biases against particular sociodemographic groups, that they minimize the impact of those biases. 
Nick Singh: 32:07
Right. And forget the HR team. You’re a data scientist, you’re a machine learning engineer. Do you want a smooth talking boss that might not know the basics of regression or can’t write a SQL query? Or do you want someone who at least has some technical basis? So that’s the other thing that they look for is at the higher levels, it’s so hard to see who’s good or not, but at least technical, you can’t lie through a technical that easily. So it helps level the playing field there. 
Jon Krohn: 32:36
Cool. Great insight, Nick. I love the conversation so far. So in addition to all of those careers that you encapsulate, so career coach, software engineer, data scientist, now product manager, in addition to that, you also used to be a DJ. 
Nick Singh: 32:54
Yes. So the thing that drives me is helping these data scientists with interviews, whether that’s content creation, whether it’s one-on-one coaching, whether I’m product managing DataLemur, whether I’m writing code for DataLemur, whether I’m writing the next edition of the book. Doesn’t matter what you call me, I’m basically just in this space of helping data scientists land their dream job and ace the interview. So I’m happy to do whatever job that looks like, whichever day it looks like, as long as I’m servicing that goal. Because that goal gives me a lot of fulfillment. People value it. People pay me for it. I have fun with it. And it’s a net good for society to match people with good jobs. I know you do some of that work at Nebula as well. And it just makes you feel good. So that’s kind of why it doesn’t matter what you call me, this is kind of the space I operate in. 
Jon Krohn: 33:47
Nice. Well, so then how does the DJing fall into that? 
Nick Singh: 33:52
DJing doesn’t, DJing, that’s just for me and my own vanity, man. I think I’ve got great music taste. I’m a huge Drake fan. And I know probably some people are rolling their eyes, like music taste, Drake fan? But no, I’m a certified lover boy. I love Drake. And I actually used to DJ Bollywood and hip-hop parties back in high school, and then a bit in SF. And DJing is really how I got my taste of entrepreneurship and doing things on my own. And it wasn’t so much the fact that I could make money by myself, it was this fact that I went to a really nerdy high school. And then I was a nerd in the nerdy high school. I used to be on the math team. I used to code. I was a nerd, I’m pretty sure. I’m kind of chubby. I was a nerd’s nerd, okay? 
 
34:39
But I also love hip-hop music, I love music in general, and I’m like, “You know what? Why can’t I be a DJ?” So I started DJing and a lot of my friends were like, “What are you doing? You’re skipping math practice for this? What are you doing?” And then I had my AP exams, I had all kinds of academic demands, but this was just something that I could do on my own. There were no right or wrong rules. And after I was able to DJ about 25 sweet sixteens, grad parties, school dances, a few engagement parties. No weddings, but people had asked, but I was like, “Yo, a wedding is way too much work. And I’m like, 17. This is too much.” But they kind of gave me my taste of first time being like, hey, people thought I couldn’t do this. Or they thought, this person’s not a DJ. 
 
35:24
This guy’s a mathlete and science nerd kind of guy. How am I going to DJ? And I was like, hey, you know what? Maybe the world’s a lot more malleable than we think, which is why even now I’m fighting to say, hey, don’t call me a career coach. Don’t call me an author. Don’t call me a product manager. I just am doing things to help data scientists, and I’ll take whatever role that takes, whatever that needs. And I think that DJing was the first place where it’s like, oh, you want me to help you with lighting? I should move my big ass speakers to the event? Oh, I need to market my DJ service? Oh, I need to have some good customer service? Oh, I actually need to know something about music and DJing and equipment and be in tune with what’s hot on the charts? 
 
36:06
I’ll do everything, because I love partying, I love putting on a good show and I love creating a great community event, and just getting a big group of people, having fun and happy. And that’s what DJing was. So I’ll look at the lighting, I’ll look at the music, I’ll do anything that it takes because that was just so much fun. And I think that’s the same thing I do for data science, which is like, I’ll do anything as long as I’m having fun with it and it’s adding value to other people and people like what I’m doing, I’ll keep doing it. 
Jon Krohn: 36:33
Nice. Do you ever DJ still today? 
Nick Singh: 36:36
COVID kind of messed things up, and something else that’s very interesting to me is I would DJ when I felt that I couldn’t tap into my creative side, in my, let’s say, businessy side. In high school, it was a lot of AP exams and SAT scores. And then in San Francisco it was a lot of code at Facebook or, hey, go work on SafeGraph. So on the side, I got to do these creative things. But Jon, these days I don’t have that creative energy. I get to spend every day making some cool meme on LinkedIn or some video or coming up with a new feature for DataLemur, and I get to act on it immediately, because it’s all mine, it’s all right in front of me. So these days it’s not just that COVID kind of messed things up, it’s just my creative energy is being fully channeled into all these projects I’m working on. 
Jon Krohn: 37:25
I get that, man. That’s cool. So like your DJ side hustle or any of the work that you’re doing now, it requires tireless self-promotion to be successful, as you’re alluding to with all these social media posts. The only way that you’re going to be successful is if you’re comfortable doing that kind of self-promotion. And for data scientists, I think especially early career data scientists, that same kind of tireless self-promotion is critical to landing their first great data science job. So do you have suggestions for listeners who maybe today aren’t comfortable with that kind of self-promotion? How can they get over the issues? 
Nick Singh: 38:07
So this is a mindset problem, so let me tell you my mindset. First of all, I don’t even think I do shameless self-promotion. I just think the things I’m building are awesome. My book is awesome. My platform’s awesome. The platform’s free. 
Jon Krohn: 38:18
I didn’t say shameless, I said tireless. 
Nick Singh: 38:21
Tireless, shameless, whatever you call it. I get that that could be seen by other people, and same way when I’m DJing, I’m like, oh, this guy’s just trying to promote his DJ thing. But it’s like, no, I love putting on parties, I love what I do. So I think that’s the first thing to realize, which is a) I’m super passionate about the work, and to get hired, you want to showcase that passion too. I think portfolio projects are a really good way to show your own passion. My portfolio projects, the things I worked on that got me hired was a little bit of DJing, or my side startup that I did in college that used data to… It was basically like fantasy football, but for musicians, where instead of drafting football players, you would be drafting hip-hop artists onto your fantasy label. And then using Spotify data we’d be able to price how good is each artist and then kind of reward people who are able to pick up and coming artists. 
 
39:15
It’s called Rap Stock, like a stock market for rappers. That was what I did in college, because it combined music, it combined data and it combined code. And I grew that to 2,000 month active users. I didn’t make any money, really, but I grew that because I loved music, I thought I had good music taste, it used data and I just had fun with it. And that’s kind of why I got to do, it’s called growth engineering at Facebook, which is again, combination of data and software engineering and marketing to grow the Facebook product, to run a lot of AB tests and understand the AB testing data, and to increase the user base and engagement of Facebook in data-driven ways with rapid experimentation, which is kind of how… You can kind of see how the music is combining with entrepreneurship, combining with marketing, combining with data and it kind of all flows together. 
40:05
So I think that the first big thing is be passionate about your work. And the best way to be passionate is, well, look, you might not be passionate about Gaussian mixture models or about this boring Titanic data set. But if in your own portfolio project, you get to pick something cool you’re interested in and you do it on your own, you have no reason not to be passionate about it when you’re talking in an interview, and that’s what gets you hired; your passion for learning and your passion for data science. Especially for an early career person, because, look, an early career person might not have all the skills. So who are you going to bet on? Two people, both need to do a lot of learning because they’re early career, but I’m going to bet on the passionate person who’s just genuinely happy to be doing this work than the other person who might be there for the money or for some other reason or just because they fell into it. 
Jon Krohn: 40:50
Yeah. I couldn’t agree more with your guidance here. And it is the number one piece of advice that I give to people when they’re asking me what they should be doing to get hired in data science or in software engineering or any other kind of challenging role. Absolutely, portfolio projects where you came up with the idea for that project, or even better, like you’re saying, an entrepreneurial side hustle, that can be more commitment, but that’s an even better… 
Nick Singh: 41:16
You don’t have to do that for a data job, but just finding a cool data set that you like, that people haven’t beat to death like Titanic or the Iris data set and then doing something interesting with it and telling a story, making interesting visual, puts you already in the top 20% of candidates. And the other part about being comfortable with promoting yourself is it starts from having confidence around your own skills. The confidence comes from the skills and the confidence also comes from knowing that those skills are yours. It’s not because you got a certification. It’s not because so-and-so said you got an A in the class. It’s because you were there, you built these projects. You can practice your questions on DataLemur and you can verify like, oh yeah, I actually do know what I do. I do know what I’m doing. And that’s where the confidence comes from and that helps you promote yourself better, because when you’re genuinely passionate about it and you’re confident in your own abilities, promoting yourself is not hard. 
Jon Krohn: 42:12
Nice, great answer. I’m so glad that I asked that question. That was a brilliant, helpful answer and I agree with everything that you said. So digging a little bit deeper into this interview prep segment of the interview, in the behavioral interview chapter of your book, Ace the Data Science Interview, you have a section about when an interviewer asks the question, “Do you have any questions for us?” So the employer, the prospective employer is asking the candidate, “Do you have any questions for us?” And so people often don’t take advantage of that moment. Some people just say, “No, I don’t really have any questions.” And that for me as an interviewer is definitely a disappointing experience. Job interviews are an opportunity, I think, not just for the interviewer to be learning about the candidate, but also for the candidate to be learning about the company that they could be working at. 
 
43:09
So these questions give you a great opportunity to see if this is the right company for you, whether the team is going to be the right fit, is the manager going to be the right fit. So these questions can be valuable for that. But I think, based on our conversation before we started recording this episode, I have a feeling that you’re going to suggest, and I totally agree with this, that asking those questions when the interviewer opens up the opportunity to ask those questions, while the candidate should have questions prepared, those questions actually should be to impress the interviewer anyway. 
Nick Singh: 43:47
A hundred percent. Look, people often say, it’s them interviewing you and you’re interviewing them, but let’s be real. You are there to try to get an offer. Even if you don’t want to work at that company, you can use that offer to negotiate against other companies and maybe they give you an amazing offer so maybe you do end up wanting to work there even if it may not have been your first choice, because they win you over later or they negotiate really well. So while you’re in that interview, you don’t want to waste time trying to figure out, is this the right company for me? You want to be impressing them and you want to be getting closer to getting an offer. Later on, through back channeling with friends and references and Glassdoor and talking to recruiters and during negotiation, you can learn more about the company and try to drive a bargain, a hard bargain. 
 
44:33
But while you’re in the interview, you got to be trying to sell yourself, and asking good questions is one great way to do that. Now obviously, you don’t suck up and be like, “Hey, why do you think I’m such a good interviewee?” Nothing like that. But just something like, “Hey, Jon, you told me at Nebula you’re working on talent matching. How do you think about underrepresented minorities and matching them? Have you noticed they’re missing some features or how do you make sure there’s not bias?” Suddenly they’re like, “Whoa, this person knows something about our company and they asked somewhat of a nuanced, interesting question,” not just, “Hey, Jon, how do you like working there?” Everyone asks that, right? So same way, you can ask something more nuanced, you can ask something more topical. You can show that you’ve done your homework. 
 
45:17
Like, “Hey, Jon. I was actually on the Nebula website and I signed up for the candidate portal and I noticed you asked me 36 questions. Do you use NLP to process those answers? How come it’s unstructured text? I’m curious. What does your NLP stack look like?” Suddenly you’re thinking there like, “Oh, wow. They actually went on our website and pretended they’re a candidate. And they actually are asking about NLP and now we can nerd out. ‘I’d rather hire that person who’s nerding out with me than the person who’s just like, “Oh, how many days off are there? How are the benefits? How do you like working there?” Figure that out later. 
Jon Krohn: 45:54
Thank you, Nick, for exploring the Nebula platform that I recently launched, which I don’t think I’ve talked about on here yet, but we just have now, in Beta for the first time, a commercial platform. So my company Nebula has now released its first product into Beta. And it’s awesome, Nick, that you explored that online. And absolutely, when interviewees come in and have thoughtful questions about the platform, the company, the more in depth, the better, huge green flags, and love having those conversations. And then on the contrary, if I ask the question, “Do you have any questions for me?” and they don’t seem to know anything about the company or they don’t seem to have interest in the company, those are huge red flags. 
Nick Singh: 46:42
Right. Especially at smaller companies where it’s not like at Facebook or Google, there’s a thousand divisions so who cares about learning about this sub team? When I’m talking to the CTO, the CDO, the founder, and I don’t have good questions for them at such a small company, it’s like, you think the founder wants to hire someone to work on their mission with them that doesn’t really know anything about it? No. So if it’s a small company it’s even more important to know the product, to know who you’re talking to and to ask something more nuanced. That’s an easy way to stand out and doesn’t require super advanced data science skills. It just requires a little bit of finesse and a little bit of prep work, which is exactly the kind of stuff we talked about in the book, and why I say data science interviewing is not the same as being a great data scientist. They’re somewhat different skills. 
Jon Krohn: 47:30
Nice. Yeah. A hundred percent. Speaking of data science skills, what tools or skills, Nick, do you think are best for data scientists getting hired? 
Nick Singh: 47:40
Yeah. Again, I love SQL and that’s why we did DataLemur, but without plugging that again, the main thing is that databases aren’t going anywhere, and a common first or second round interview is SQL. Before you even get to talk to a person and get to impress them with your AB testing skills or your MLOps skills, there’s no MLOps first round interview. It’s usually a SQL interview or a Python coding interview. So those are the two places to really focus on, Python and SQL, when it comes to just getting hired. And then I think the other tool, well, it’s not really a tool, is I think it’s good to focus on the things that aren’t going out of date. Because Python will come and go, SQL may come and go, or different flavors and syntaxes might come and go. But just this idea of, hey, how do we store data well? 
 
48:31
How do we structure tables or organize data better? How do we draw insights from them better? How do I communicate those insights better? Those are the things that I think that people should be focusing on and leveling up the most, because I can tell you, 30 years from now, we’re still going to be asking and answering the questions of how do we organize data? How do we analyze it? And then how do I communicate those results to people who might not be as technical as me? So that’s what I find the most interesting thing for people to focus on because learning PostgreSQL 14 or mySQL server, whatever version they’re on, meh, it’s okay, but… 
Jon Krohn: 49:06
Yeah. That’s a perfect answer. So obviously SQL is going to be a big thing, given everything else that we’ve talked about in this episode so far. And as I’ve mentioned on the show in the past, SQL is the most popular tool in data science, period. So you’ve got some people that are Python users, some people that are R users, but almost everyone in a hands-on data science job at some point needs to be making SQL query. So a hundred percent. And then I love that, the focus on timeless skills around organizing, analyzing, and communication and communicating data. That is so true because it is timeless, and so investing in being able to speak about that knowledgeably is going to be useful for you now and in 30 years. Nick, this has been a fantastic episode. I’m sure that our listeners have learned a ton. But as we start to wrap things up here, do you have a book recommendation for listeners, other than, of course, your own Ace the Data Science Interview book? 
Nick Singh: 50:03
Yes. I love this book called Antifragile by Nassim Taleb. The writing style is kind of grating. He insults everybody and everything. But he speaks a lot of truth and I found it to be very interesting, and that book has changed my diet, it’s changed my career trajectory, it’s changed how I approach a lot of different situations. And it’s a book that’s very hard to summarize because it’s about risk taking and probability distributions. And it just shows up everywhere and he just keeps applying it to all kinds of stuff. 
 
50:38
So again, it’s a really good book for data people, and there are some mathy things in it, but it’ll talk about politics, it’ll talk about diet, it’ll talk about how doctors might not be giving you the best advice. And it’ll talk about career risk, and I’m building an anti-fragile career, which is, you can’t fire me. Maybe you could cancel me because I’m in the public eye, but you can’t really fire me. Recessions don’t really hit. I had mentioned my book has been being counterfeit, but I have other business lines I’m working on, DataLemur one of them. So I’m just building a very resilient portfolio of bets rather than putting all my eggs in one basket. 
Jon Krohn: 51:16
Brilliant. That sounds like a great recommendation. I would love to make the time to read that. Awesome. And then, so Nick, you are a tremendous content creator. You have nearly 100,000 followers on LinkedIn at the time of recording. I’m sure that by the time this episode is released, you will have six figures in your followers there. So I can imagine following you on LinkedIn is something that our listeners should be thinking about doing. Where else can people follow your work? 
Nick Singh: 51:46
Yeah, check me out on Twitter. Again, Nick Singh Tech. And I also have a newsletter where I share a bunch of different career advice that’s not just for data folks, but just how to think about careers in general. And then of course, go check out DataLemur and the book, and you’ll find plenty of more information there and ways to stay in touch with me there too. 
Jon Krohn: 52:05
Nice. We’ll be sure to include all of the links to all of those resources in the show notes for the episode. Go check them out. Nick, this has been an amazing episode. Thank you so much for your no-nonsense practical advice on acing the data science interview for all of our listeners. And we’re going to have to have you on the show again sometime in the future to give us an update on all the latest in the data science interview space. 
Nick Singh: 52:27
For sure, Jon. Thank you so much for having me.
 
Jon Krohn: 52:29
What a guy. I loved Nick’s energy and had a blast recording this Super Data Science episode with him. In the episode, Nick filled us in on how his DataLemur platform was a natural extension of his Ace the Data Science Interview book, because it provides an instant free place for people to practice SQL queries with real world data and examples. He talked about the three major misconceptions about data science interviews at the tech giants, that is, more is not necessarily better, certifications are not a big deal, and even senior roles will require you to demonstrate your mastery of the essential data science skills like writing SQL queries. He talked about how portfolio projects and entrepreneurial side hustles are great for developing your self-promotion skills and landing your dream job, how you should always try to an offer in every role you interview for, because even if you decide the firm isn’t for you, you can leverage the offer during other interviews. 
 
53:24
And he filled this in on how SQL and timeless skills like organizing data, analyzing data and communication are pivotal to landing the best data science roles, both now and in the years to come. As always, you can get all those show notes, including the transcript for this episode, the video recording and any materials mentioned on the show, the URLs for Nick’s social media profiles, as well as my own social media profiles at www.superdatascience.com/615. That’s www.superdatascience.com/615. 
 
53:50
Thanks to my colleagues at Nebula for supporting me while I create content like this Super Data Science episode for you. And thanks, of course, to Ivana, Mario, Natalie, Serg, Sylvia, Zara and Kirill on the Super Data Science team for producing another invaluable episode for us today. For enabling this super team to create this free podcast for you, we are deeply grateful to our sponsors. Please consider supporting the show by checking out our sponsors’ links, which you could find in the show notes. And if you yourself are interested in sponsoring an episode, you can find our contact details in the show notes as well, or you can make your way to jonkrohn.com/podcast. Last but not least, thanks to you for listening all the way to the end of the show. Until next time, my friend, keep on rocking it out there and I’m looking forward to enjoying another round of the Super Data Science Podcast with you very soon. 
Show All

Share on

Related Podcasts