DataFramed
DataFramed

Episode 90 · 7 months ago

[DataFramed Careers Series #4]: Acing the Data Science Interview

ABOUT THIS EPISODE

Today marks the last episode of our four-part DataFramed Careers Series on breaking into a data career. We’ve heard from Sadie St Lawrence, Nick Singh, and Khuyen Tran on best practices to adopt to help you land a data science interview. But what about the interview itself? Today’s guest, Jay Feng, joins the show to break down all the most important things you need to know about interviewing for data science roles. Jay is the co-founder of Interview Query, which helps data scientists, machine learning engineers, and other data professionals prepare for their dream jobs.

Throughout the episode, we discuss

  • The anatomy of data science interviews
  • Biggest misconceptions and mistakes candidates make during interviews
  • The importance of showcasing communication ability, business acumen, and technical intuition in the interview
  • How to negotiate for the best salary possible

[Announcement] Join us for DataCamp Radar, our digital summit on June 23rd. During this summit, a variety of experts from different backgrounds will be discussing everything related to the future of careers in data. Whether you're recruiting for data roles or looking to build a career in data, there’s definitely something for you. Seats are limited, and registration is free, so secure your spot today on https://events.datacamp.com/radar/

You're listening to data framed, the podcast by data camp. In this show, you'll hear all the latest trends and insights in data science. Whether you're just getting started in your data career or you're a data leader looking to scale data driven decisions in your organization, join us for indepth discussions with data and analytics leaders at the forefront of the data revolution. Let's dive right in. Hello everyone, this is a dull data science educator and evangelist at data camp. Welcome to the last day of our four day data frame career series, where we deep dive into the INS and outs of launching and building a career in data I hope you've been enjoying the series so far and I'd love for you to let us know what you've thought about it, so make sure to tag us and let us know what you think. Throughout the series so far, we've really focused in on how to make it to the data science interview, whether it's through building a personal brand, which we discussed on yesterday's episode with Queen A, seeing the data science Portfoyo, which we discussed on Tuesday's episode with Nick Sing, or optimizing your resume, which we discussed on the first episode of the series with sad st Lawrence. However, there is no building a career in data without actually solving or acing the interview. This why I'm so excited about today's guest. Jay Fang, is the CEO of interview query, a remote data science interview preparation platform, WHO's on a mission to help every data scientists land the job. You can think of it as the lead code of data science, and Jay has helped tons of data scientists prepare for technical interviews. Throughout our conversation, we discussed the anatomy of a data science interview, the biggest mistakes candidates make during interviews, how to avoid them, and the importance of showcasing communication skills and business acumen during an interview, and much more. If you enjoy this episode or the series, make sure to rate the podcast and leave a review. I'd love to hear from you. Also, just to Sadie and nick will be delivering talks on June twenty three data camp rate our summit. Jay will also deliver a workshop on a seeing the technical interview, so make sure to register for the event by going to events dough data campcoma radar. The link is in the description and registration is free. Now on today's episode. So I am super excited to talk to you about best practices for acing the data science interview and all that fun stuff. But before can you give us a bit of a back run about yourself and how you got into the data space? Yeah, so I started out as a Undergrad at you dub setting electrical engineering, eventually realize I didn't like it that much, got tinkering around with like the circuit boards and the soloscopes and realize, you know, I could probably do something more interesting with my time and eventually pivoted to data science by just doing a lot of self learning online. I got into data science through kind of like the rise in two thousand and fourteenwo thousand and fifteen with pandas and being able to like jump in on Python and analyze all this data and eventually got a marketing analytics job in Silicon Valley when I graduated. And Yeah, that kind of spearheaded the...

...whole career into actually becoming a data scientist and then eventually starting my own business in the data science space as well. That's awesome. And now you run. Can you spend it on your business as well. Yeah, so I'm running interview query. So interview quarry is a data science interview prep platform and so we're kind of known as like the leak code for data science. Basically, anything you need to know about the data science technical interview, whether it's around key studies, specifically programming, sequel, take them assignments. We kind of put in a lot of practice problems on that and then get you situated into understanding exactly how to tackle and learn these problems from like a first principles kind of thinking. That's really great. So you're the perfect guess. That covered today the data science interview and I want to cover all aspects of the data science interview with you today. So, of course data science has matured over the past few years or so. Companies now have a much better idea about what makes a successful data scientists today than they did a few years back and they try to filter out for these skills in the interview process. So, given that I want to start off with the first question, I'd love to first, I'm back, what the anatomy of a data science interview looks like today and kind of what are my main skills? Hopeful applicants need to showcase throughout the interview process. Yeah, so data science interviews today are pretty wide across a range of topics and I think that's what makes them actually pretty difficult. Like, for software engineering interviews, you can generally grind, leak code and like, you know, study how to basically tackle these programming type, algorithm based questions that are tricky, but you know, after time they all look the same, right, and for data science interviews it's very much different. You'll get some probability questions about coin flipping, you'll get statistics problems about distributions. You might get even like a business case study asking you about golf balls and airplane or something like that. Right. So the difference across it is what makes it difficult. And so a lot of it is about, you know, cementing and understanding exactly which topics are going to be actually important for the interview and are going to come up and specifically what kind of roles your like specifically looking for as well, so that you can cement the specific seals into those roles. And so, for example, analytics roles are going to test you a lot more on sequel and dating, analytical thinking as well. All those kind of like some product case studies, whereas machine learning and like more modeling data scientist roles are going to test you a little bit more on machine learning concepts and programming, and you know how you deploy systems as well, and so I think being able to understand those specific skill sets really well and understand kind of like you know, what your strains are, is really helpful for getting through the Navy process. That's really great that. I'm excited to unpack these different topics, but before I kind of want to also ground the discussion on the interview process itself. So, giving your experience helping folks through the interview process, what do you think is the big misconception people have when it comes to the data science interview? I think bigness consented...

...question is the focus on only technical skills and I think there's this gray area between, you know, just straight up behavioral questions where they're asking you about culture fit and if you like working with other teammates, versus, you know, strong technical skills. There's all these like ride range of questions in between that people don't really expect when they get on the interview and then they have to perform well On for example, something that's a little bit more in between. It is just talking about maybe like pasty to science projects or defending your resume when you write about your like past stay to science projects as well, right, and I think a lot of people kind of expect that. You know, if I do this one project on Caggle and I follow the statorial and it evolved deep learning, than I can put it on my resume and no one's going to ask me about it. But you know that's not entirely true. Right, and I think a lot of the times, you know, if you put a lot of buzz words on your resume, they're going to expect you to really understand the fundamentals behind that. And that comes into like this kind of gray area between is it behavial and of your question? Is it kind of just talking about your previous experience? What about kind of like case questions where they're asking you about like how would you think about, you know, this business kind of strategy as it relates to data science? A lot of those questions are really hard. People don't really expect them and but they're really important for data science roles because you do and that a lot of your judgment in the job and in the interview, and people expect you to have like, you know, twenty percent better skills initial stakes, but also really, really smart kind of judgment, and so I think those things are things that most people don't really think about when it comes to they you process. They think they can just kind of like grind problems and then kind of repeat them when it comes to the interview. But may be surprised at how much you know, communicational, cross functional work is involved in a data science role and how companies also test that. So and some are do you think kind of the skill set that you're describing this gray zone? It's kind of combination of business acumen and data story selling skills. Is that correct? Yeah, and I think it's specific towards the role of like an analytics kind of data scientist. I would say that, you know, if you're going for more a machine learning role or one where you're like very technical, you're just working in code, you don't really have any insight into maybe like the business operation except for the specific requirements that's going into your model, then you might not actually have to deal with this because you know they're expecting you to like do a much better job at deploying machine learning models at scale. But if you're thinking more so of like kind of like this product manager that has like data technical skills, as like a data scientist, you know, it's like you don't have to go to someone else to ask them to analyze this data, you can analyze it yourself and come up with a conclusion, then I think that's where you're really expected to know these business skills and these kind of analytical skills to get you through the interview. That's really great. So let's start off with the technical interview. I'm excited them back with you kind of all the types of technical interviews and topics that you can have in a data science interview. Arguably this is one of the most discussed and arguably the...

...most the interview practitioners are worried about the most. So can you walk us through the different types of technical interviews out there for a data science role in the type of topics that you may encounter in a technical interview? Yeah, so, if we talked about like the different kinds of technical topics, I can go through probably like ten different ones if we want to go through them all. But specifically, I think, or as starts out is there's more like the case steady ones, right. That's more like product metrics, data analytics and business case, and so these are more so questions that will ask you about investigating different anomalies, understanding success, how would you like kind of Ab test of feature? Those kinds of questions are very much focused around this kind of you know, product manager, data scientists, data analytics, hybrid and those are, you know, I think, one of the ones that we focus on a lot at interview query because they're kind of like the sneaky ones that you don't really know if you got the right answer. Not. Arguably, there really isn't and right answer. If I ask you like how would you measure the success of the data frame podcast, right, we could probably talk for a really long time about all these different ways that data frame podcast contributes towards data camps, overarching goals and mission. Right, and so these are the ones that I think are the most tricky for most data scientists. Then we kind of get into more of the technical case study stuff, and so that's more so we're talking about machine learning concepts atentionally, even deep learning. This is even like probability statistics. These types of questions all kind of go around this mathematical and conceptual idea, around like how much do you know machine learning right? And so this could be, like you have a model that you want to build for, let's say, predicting the amount of, you know, subscribers are going to get on a podcast. How would you build this model? Like which kind of model? What you use? You know, like how would you evaluate your model? And so it's a kind of like a full walkthrough. I like how much you know about machine learning when it contributes to like an actual business versus you know, a model and it contributes to a very specific kind of like technical you know, just case study around it. Like it's not going to be the titanic data set, it's going to be something that the company has to deal with this. Well, the last round of kind of like en of your questions is then programming questions, sequel questions, data cleaning questions, and so these are all around kind of like how quickly are your actual core technical skills right? If I ask you to like draft this report or get this metric, like how quickly can you do that? Are you going to take a lot of time and are you going to mess it up. Are you going to give me data quality issues or are you like really, really sound technically and can you like, you know, apply these like theories, set theory, you know, in practice, and really retrieve the data and I'm looking for? And so that's probably like the last kind of grouping of technicality skill. That's really great. Thank kind of given the type breadth of technical topics, if available,...

...you mentioned here at least like three big categories of different subscles that you can have and the different skills data roles from different companies to companies. How should applicants prioritize which technical tool pics they should prepare for and their interview process? Yeah, so I think one big thing that an applicant should first do is just go through the role description and understand exactly what kind of role it is. Right, like we kind of talked about like the two to three different kinds of data science roles, and so just understanding. You know, am I under like this product and analytics team, or am I under this like more machine learning heavy kind of team, or is this like a startup where I'm doing everything right? And then I think it becomes more understanding the interview experiences previously given at that company. So if you can read through and understand, you know, what previous interview experiences other candidates have gone through, you can kind of get any understanding of like just what actual data science rules and skills are going to be asked. We do this thing in your quarry to where we've created these guides and we've really worked on making these guides good in terms of understanding. We scrape a lot of interview experiences, we kind of do some classification and we just kind of show a radar chart of exactly, you know, what skills are going to come up for each interview, just so applicants and good idea of which question types they need to actually practice as well. And then, you know, the last thing you do is just to make sure you can bench mark your skill set across these different question types. So if you know you know that they're going to ask a lot of sequel questions, then definitely practice like a ton of sequel questions. See how good you are at sequel. If you can figure out your in like you can pass, you know, like some easy, medium hard questions on the code or any of your query and you know the questions the ass that they interview are going to be around like easy to medium, then you should be good to go right so just really understand, you know, your skill set and which which ones you need to study for the most. And if a company, let's say, doesn't have a lot of kind of best practices or a lot of available online resources on what the interview persons would look like, what will be your advice for someone kind of looking to understand and prioritize which direction they should go for in preparation for a particular interview? Best Way to overcome that is just, honestly, the talk to the recruiter or hiring manager and ask some you know, what they test for, and just say like something like I want to know that this is a good fit for my skill set. Could you please, you know, go over like what kind of role this is or what might be on the interview so that I can be well prepared? And that should be like pretty good enough. It's very understanding that you might want to know what's going to be asked. I mean I think most people know that like these interviews are pretty random across like potentially ten different topics, and just asking that question shouldn't hurt. That's great. So you've helped a lot of people kind of go through the interview process. I wonder from you, from your perspective, what do you think are the most common mistakes applicants make during the technical interview? So the most common mistake that I've seen is that candidates generally will get tripped up and jump right into a question when they should probably really take a second to think about it first.

I think there's this idea around measuring your success in an interview across three different specific areas, right. One is kind of business acumen, another is communication and the last one is like technical intuition, right, and so across those three, all the time people kind of jump right into like the technical part. They like you ask them to build a model, and they jump right in into like the technical details around, you know, random forest or something like that, when they should actually just be getting any more information to understand the actual situation that they're in. A lot of the Times you're in kind of like this information disadvantage when you're working in an interview. Right you're in this like forty five minute close time slot where basically the interviewer is asking you about a problem that you've never heard about or a situation that you've never even thought about in your entire life and then you suddenly have to kind of like come up with an answer that's just as good as if they were maybe talking to a CO worker with that kind of disadvantaged you know, you really have to get a lot of questions, you have to get a lot of like answers and a lot of context really quickly, and so I would always advise to you know, ask questions, take some time to really deeply think about the problem and then structure your answer in a way that's kind of cohesive and allows it for you to stay a lot of assumptions up front so that you don't make some you know, really kind of dumb the stake by just diving into an area that is completely different than what they expected. I complete agree. Especially jumping into the technical part, you miss out on a lot of context that could help shape even the type of technical solution that you want to provide within the interview process. So can you walk is through kind of an example of this common mistake that you're talking about from real interviews that you've seen and what would have been a better alternative? Yeah, so, for example, there's one common question at facebook that's something like, you know, facebook is seeing that parents are joining the platform to and adding like their teenagers as family members, and the APP right and how exactly, if you measure, like the effect this has on teenagers? And a lot of people will jump directly into just saying about how they would like build a model to basically see the teenagers turning off of the platform, and that's not really like the right way to bring it right, because a lot of the times I would say initially, like how do you even know that the parents and the teenagers are linked up, like how do you even have that data to understand that this parent is like a parent of this teenager? How do you know that their family members like, how do you know like the age of like these different users? And so just like clarifying a few things, understanding your data set and then deriving some causal, you know, relationships afterwards is probably a better bet. But yeah, I think it really does depend on like how you think about the problem and I think you know, sometimes if you can like successfully argue your point, then that's should be fine. You know,...

...if you're extremely technical person and they know that they're hiring you for F for extremely technical role, and they should be okay if you kind of jump into like this deep learning model or something like that, but otherwise for everyone else is kind of like tackling these questions. You really want to kind of apply that judgment and be a little bit more forward with how you approach things. So, given that, what would be your recommendations kind of for a study plan to be able to show up some of these gaps technically minded people may have that they don't necessarily have in the same sense round ten of this product sense or being able to approach problems constructively. Yeah, so I would say that the best thing you can do is to benchmark kind of your skill set, as I said before, and just understand like what level you have to be at to do well. The best way to do this that we've seen is to actually conduct kind of like a coaching session with someone who's like seen a lot of people that has passed the interview or hasn't passed the interview, and they would then kind of understand the level of what you need to be at. And so if you can kind of do an initial mock interview on like a couple different question topics and get like that benchmark, especially for those questions that are a little bit more open ended, then you can like understand basically how much you know content you probably need to then study to then get to that level. You know, if they give you, you know, like three out of ten rating basically, and this is what we do Intervi Querry as well, we basically assigned kind of like a score after the end of the mocking your view, then you know that like basically you have to probably do like x amount of more setting and so generally for us, you know, like if you're at the fifty percent mark, you probably need to study, you know, like for another three the four weeks, right, but if you're already at like the eighty nine percentile in terms of like how well you're doing on like a particular question they asked, then you probably just don't need to say that much. You probably only do like one to two more questions, and so it's really about understanding just like your level when you're given different like problem types. Okay, that's perfect. So let's jump over from the technical interview to discuss more kind of the behavioral interview, data storytelling, business interview. This running extent, arguably one of the most important aspects of the interview process is showing your ability to communicate. We touched on this earlier, but I'd love to expand on this by getting from you just how important is this skill from a hiring manager's perspective and why I think it's super important. I think a lot of the Times I've actually heard that this is like the biggest red flag that they get from candidates. A lot of them will be able to pass like the sequel interview, but then when it comes to just behavioral questions I'll stumble a little bit. And the reasoning before that is kind of like what we talked about kind of display and judgment and being able to communicate your finding. This kind of like thing about like if you can't communicate your finding as well, did you even analyze the problem right? Is kind of like if a tree falls, did anyone hear it? Sort of question, right, it doesn't really matter if you do a really good job of an analysis if no one can really understand it. And so a...

...lot of the time I think hiring managers really want to see that like ability to communicate insights and be able to craft that kind of data storytelling ability so that other people understand what you're trying to convey. And especially at bigger organizations, you have a lot of different people at these companies, right you might have to convince like hundreds, maybe in thousands, of people to like actually like read your analysis, maybe take something away from it and understand something. To do that requires, you know, pretty good skills in terms of either writing or communication, and kind of keeping that in mind, I think that's probably why it's such an important skill for, you know, all these kind of technical roles. And I guess even if you're on like a very technical role, like a machine learning engineering, deeply research scientist role, you still also have to like understand exactly how to communicate technical concepts within your team and work together as well. And I completely agree with that notion, because gaining adoption as a data scientist right is what makes her breaks kind of your work down the line once you work in a company. So, as you said, your analysis doesn't really matter if no one's being able to consume it or if you're not able to kind of create decisions or generate decisions as a consequence of it. Exactly, and I think that this is pretty underrated. I think also increasingly, as we become kind of integrated into a remote first world, this is going to be even more important, right in terms of just like written communication, and this is going to be a lot more widespread, I think, in terms of how it affects just general like workplace culture. So getting a leg up on that. Having good, you know, writing a communication skills is always like pretty necessary. I think for any technical that's great. So, similar to my previous question on the technical interview, what do you think are the most common mistakes applicants make when showcasing their business acumen in communication skills and an interview process? The biggest mistakes, I think, and I think I've kind of touched on this too, is just kind of jumping right into it, I would say, additional things that generally could be worked on. is also kind of like conversationally structuring your answer so that it flows well. Lot of people have this kind of jumbled thought. I'm definitely one of those two and which, you know, if I think of an idea, I'll kind of jump in and say it, but I won't really think about how it relates towards like the overall structure and you know, and writing. We have this whole like concept of and like an Intro, a body and the conclusion, but that's really hard to do when you're just, you know, speaking to someone conversationally and you've like creating this pitch out of nowhere. So just adding a bit more like, you know, structure towards like you're thinking definitely goes a long ways towards the other person understanding it a little bit better. I completely rae, especially like on that kind of structuring your conversational thought. It's definitely a difficult skill to master and, given that communication skills are inherently not as straightforward to improve as technical skills, how do you recommend people to kind of shore up these gaps, and especially as they go through the interview process? Definitely the number one thing you can do is...

...just to do more mock interviews, whether that's with peers, co workers or even like professional coaches online. Just getting in that repetition format of basically repeating like the same situation of the interview, but, you know, without as many stakes, is like a great way to like improve and I think just like doing that again and again it's like probably the best way to tackle it, because this is kind of like a weird foreign content. Usually people don't ask you for like your technical input on like a question that you've never heard of, right and ask you to respond within like twenty minutes on like how you would approach it. And so you're doing a lot of like deep thinking and you're also having to like respond to it and then structure your thoughts and then also like make sure that it makes sense, and then you have to jump into like another kind of realm if the interviewer ask you a different question. And so, overall, I would say just doing more practice. So that is the best way that you can, you know, get better. I mean, I mean you could also just say like talking and socializing with people more, but honestly, it doesn't really help with that specific skill of like deeply thinking about a subject and then also having to then craft like a solution to it. That's great. And in terms of kind of you want to break down that skill, what do you think is more important today, like writing and speaking or being able to dealver presentations? Like how would you break down kind of communication skills and did sub components? I would say writing is probably the most important this day and age, just because I think everything kind of starts from writing just to begin with, like you kind of have to start writing down your thoughts and structuring them. When you're doing an interview, it's definitely helpful. Even if you do a presentation, you're also writing down your thoughts and structuring them beforehand, at least I'd hope, so before, you know, kind of jumping in and talking about this like long graph or like a bunch of different insights that you've had. And so I would say that writing is probably the most important. It's also the most beneficial, to be honest, in this day and age when everything the synchronous and you know, for me, like, I found it to be a really helpful I can just post on Linkedin, I can write blog posts, people read it throughout the years and then come to visit our website and kind of like Wanta, then kind of read more of our content. And so for me it's been like the core crux still like building my whole business on Inter you query, and has also helped me throughout my career terms of just making sure that, like anyone who with some like sort of power or sway at responsibility at like an organization such as like the CEO or the executive team will be able to see that I have like insights to contribute right and especially, I think, with writing it's just a lot easier for you to kind of just communicate those technical thoughts and save them somewhere for you to like look back on as well. I could an agree more, especially kind of when you think about the importance of a portfolio project and how important are kind of portfolio projects getting your foot through the door through the interview process, for example, for able to showcase great storytelling skills throughout your portfolios, throughout your written content,...

...you have a much better chance of making it past the finish line than other candidates. Is that would you say? That's correct. I think so. I think just doing the portfolio project gives you that practice for it's becoming a better data scientist as well. I would say that portfolio projects are a little bit hard to get in front of people's eyes unless they have some sort of like additional kind of new insights or if you can market it really well, because of the fact it is really hard to like for someone to just look at your resume and actually like analyze it versus, you know, just looking at your resume for two seconds and then kind of moving on the next one. And so I think like portfolio projects are always good, and the main reason for that in personally, is just through the practice and the repetition of actually doing more data science. But additionally, you know, as you said, for you to actually communicate those insights into like writing and have it save somewhere. That's also like the second step that you really need to do for these projects to like make them last and also like build that repetition feedback loop in your head. That's awesome. So we covered the technical interview, we covered the behavioral interview. Finally, had love to dedicate kind of this last section of our podcast just chatting about the general best practices throughout the interview process itself. Let's say I made it through the final round. I'm having a final fit conversation. What are questions I should be asking as an aspiring data scientists know whether this role is a right fit for me? I would say that there are these general like values that a company has and then there's also like specific values that like a data science team might have, and so just understanding what those values are for yourself and seeing if those exists of the company and within the team is a great way to kind of like get started. There's this great website called key valuescom and the founder is like kind of made this into like a job board, unintentionally, but essentially you can go there and you can see kind of like company values of what companies kind of like choose in terms of like great work culture, work life balance or like good code, and you can kind of like choose your own from there and then you can find like the companies that actually have those values. So I love that website. I always recommend it and I think that those are like great kind of like lists of items that you can go and ask companies about if they practice basically in the interview as well. That's really great and find, I think, one sticking point throughout the final fit and like once you received offer and you have an all for conversation, is salary negotiations. I think this is something especially junior candidates struggle with. You mind walking through how you'd approach tell your negotiations, how you've consulted other interviewees that you've worked with to approach seller negotiations? Yeah, so definitely. I think the biggest thing is understanding your own market value out there. So generally, companies will hire you for the price that it costs to replace you. That's kind of like the brutal nature of hiring in general, and so for a candidate, how you can take that? Your advantage is to understand, you know, what your market value actually is, and so we actually just released this kind of salary report on in a requerry where we you can go in, you can type...

...in kind of your location, your role, maybe even a company, and kind of see what the average salary is across a bunch of different ranges, and I think that will really help towards understanding exactly like what you're probably going to be expecting for this specific rule in salary negotiation. In terms of actual like salary negotiaation tactics, I would say that one thing that you could do is also just to delay, you know, the conversation around salary until as late as possible, just because it doesn't really help you when you talk about it earlier in the process. Mainly I think it actually kind of hurts you because you're giving them information up front. One thing I've learned is that you can always ask about salary ranges, especially with new laws kind of coming up for a position. And then, lastly, one thing that you can also do is kind of negotiate on different things that are not really a kind to just salary. And so there's base salary, there's stocks, there's rsus, there's like Pteo, there's, for one k matching, there's all these things that you don't think you can negotiate on, but you actually can, and I think you know, I think it's something that's available and will help with like the negotiation process. That's really great. Then, harping on that salary inside report that you guys released, what were some of the key in sites that surprised you, at least once the insides were distillt? Yeah, the key in sites that surprised me the most and personally, it's just like a stack geek. I really like the fact that one, there's all these cities that people can move to that basically have a better roy on their salary than like the top cities you think of. So you if you think of like San Francisco, Bay area, Seattle, right, they're going to pay you probably the most. But if you move the cities like Boise, Salt Lake City, a lot of cities in Texas, you actually get better like Roy on your money. From a cost of living perspective. These cities are all like kind of great upcoming cities, probably better like also real estate investment opportunities if you're going to move there, but also just in general with the remote work. I think they've like grown a lot in popularity on. The second big insight was a fact that data science salaries on average have got up around like twenty five percent in the past five years, which is pretty incredible because compared to software engineering salaries, I've only got up by like around six to like eight percent, and so it's a little bit harder if you try to measure a total compensation. But specifically, I think the biggest data science roles I've seen like the most growth have been data analytics and date engineering, and both of these, like rules, have kind of skyrocketed up and like gone up by about thirty five percent in the past five years. And so date engineering, I think before is kind of like misleading because a lot of the Times did engineers are actually software engineers and softeagn nengineers. I dated engineers and they just don't, you know, distinguish them on rolls or on, you know, linked in, so that you could really like analyze it did until but this role date engineering is blowing up right and people are paying more for it and they really really care about getting good data engineers because, as we...

...all know, you need good did engineers to really enable good data scientists as well. That's really great, then. I love that first point, especially on like the different cities you've lived in. We've definitely been excited as well around the remote revolution and what that means to kind of leveling the plane field one when it temps to talent search and kind of breaking into data science, but also getting like the best type of compensation you can get for where you are now. On a final no, Jay, let's say you are data scientists who just joined a company. How do you provide value and make a good impression in your first ninety days? I always tell everyone like in your first ninety days, from what I've learned, basically just go around and meet every single person that you think is important and then ask them who you should meet basically, and then just keep on meeting people. Right. So you just basically go around kind of introd everyone that you know then definitely just try to add some business value like extremely quickly, and this is like not very difficult because there's a lot of small things that people just overlook and just don't do. And so one thing could just be like documentation, like just going in and documenting something that you're finding out isn't documented on your onboarding that should probably be documented, and people like that's huge value right, because honestly, this is just something that people just don't do and just needs to be done in the on boarding process. Then the second thing, I think, is just kind of just figuring out incremental ways to like continue to add value on top of that, maybe even from a technic perspective. So personally, for me like kind of like doing some writing during some and now assist kind of setting that out to team members afterwards. For me, like your first experience is good, and then also just contributing like small pain points where people feel like, you know, they either need like a script for something or they, let's say that they like need, yeah, and Etl job somewhere. Just like kind of figuring out exactly like how you can most provide like these small kind of like value add activities is like a great way to kind of start out and then, I think as that kind of ball starts to rule, then a lot more things will be a lot more apparent and then you'll kind of take on bigger projects. But I think it's really important to just have a few like success kind of like wins under your belt so that you can be known as someone who, like, you know, does stuff. Basically, I think sometimes, you know, you might struggle as a junior candidate where you try to like take on this huge project right when you get started, and I wouldn't recommend on that. It's more important to kind of learn a little bit of the ropes and kind of like get some wins on your belt first. I could a grade. That kind of low hanging fruit approach to impressing on your first sixty days is super important, especially in smaller organizations where kind of the time to value is a bit shorter than like larger organizations is running extent, where you're expected to have a like a longer on boarding time, and that, I think, what is a framework has helped me maneuver my career quite a lot as well. How do you kind of systematize choosing which like which problems to tackle within your first thirty sixty days. You know, the information disadvantage, as you said, is is huge. It's like basically the time to value is really, really high, and so...

...the kind of approach that you could take maybe is kind of like list down all the problems that you see, kind of assess like the scope of each one and then, as you said, kind of tackle the low hanging fruit first, right, and ideally, you know, like maybe it won't be like a huge value add to someone or highest priority, but if it's like an easy, one day, few hours kind of win, then like definitely notch that under your belt and like just prove that you can, you know, add value to this organization and I think just like slowly moving up in terms of actual projects. After you were looking at you know, that whole ist is probably the best thing you can do. Right. They're not expecting you to lead a project like thirty days out right, and you probably shouldn't be doing something like that to begin with, because you probably don't know a lot about the organization and like a lot about the business, and so it's all about like kind of slowly kind of tackling projects and like, I think, gaining that success into your belt for sure, that's really great. Finally, Jay, I had an awesome time chatting. Do you have any final words before we rub up today's episode? Not Really, I would say, for everyone who is interviewing out there, I definitely know that the mental tolls very high. This something we do with a lot interview queries that we like hear about people. You know, you get an interview, you get excited, you like go all the way to the on site and then you fail and then it's crushing, right because you're just continuously kind of getting hammered in these like interviews from like a mental perspective and then also from like the fact that of you're getting a lot of rejections. And so I say it always like stick in there. You know, it's might seem like it's like really bad, but actually all these rejections on the interview is actually just, you know, kind of like, you know, more and more practice, so you could say, towards getting more and more offers at that point later down the line. And it's definitely a grind. I would hope that, you know, you can find some ways to make it a little bit better. I personally do a lot of things where I try to concentrate my focus around things I do enjoy. I figure how I can incorporate that into like something that is a grind, like interviewing, and I think that like works well personally for me. So yeah, I would say just trying to integrate it, don't be too hard on yourself, and then definitely realize that you'll eventually throughout the grind. There is like an opening at the other end where you will eventually find a job. That's really awesome. Thank you so much, day for coming on the PODCAST. Yeah, thanks, Adel. Let's talks in. You've been listening to data framed, a podcast by data camp. Keep connected with us by subscribing to the show in your favorite podcast player. Please give us a rating, leave a comment and share episodes you love. That helps us keep delivering insights into all things data. Thanks for listening. Until next time,.

In-Stream Audio Search

NEW

Search across all episodes within this podcast

Episodes (121)