Crack the Google Data Scientist Interview | Former Google Data Scientist | DataInterview

one of the most renowned lucrative positions in data science is becoming a data scientist at google there are three major benefits of working at google as a data scientist people projects and perps when you get to work at google you get to work with perhaps one of the most bright and talented individuals out there who are very supportive in your career growth you also get to work on projects that have real impacts on millions and billions of users across the globe and lastly you have one of the best perks out there from weight room to meditation room to free food and so many other perks so do you aspire to become a data scientist at google hi i'm dan ex google and paypal data scientist in this video i'd like to discuss how to crack google's data scientist interview there are five sections i'd like to cover in this video rules interview process question types prep tips and a simple interview question if you like this type of prep content please hit the like button subscribe and check out datantv.com data interview contains courses covering topics like a b testing business and product cases and product sql and you also get access to mock interview videos and you also get access to slack community group where you can partner up with community members to practice for data scientist interviews and to get the extra boost you need in order to ace the data sentence interviews you can pair up with an instructor myself through the one-on-one mock interview coaching where i provide personalized coaching in the form of interview questions along with assessments that comes with solutions and feedback that are important in order to help you improve data science interviews so make sure you check out dataentry.com now without further ado let's dive into how to crack the google's data scientists interview starting with the role itself so when you're a google data scientist you're plugged into a specific product area from google search youtube maps gmails and android and then there are also horizontal roles such as operational data scientists and finance data scientists which will support various product areas as i just mentioned and you get to work on some pretty cool stuff from running a b experiments statistical analysis and modeling and you also get to work with various technical and non-technical individuals from analysts engineers product managers and business stakeholders and lastly the work that you do have implications on millions and billions of users which are really cool so let's take a look at the qualification itself in order to show you the qualification for this role i pulled up the job post on careers.google.com so in the job post for data scientist engineering position and it is engineering because data science is part of the engineering org within google and the minimum qualification for this role is that it requires a master's degree in quantitative discipline ranging from statistics operation research biometrics and so forth and it requires at least two years of work experience and data analysis build and you need to have experience in statistical software it could be r python matlab or pandas and knowledge and database languages like sql so this is the minimum qualification but what about preferred qualifications so the preferred qualification is having a phd in a quantitative discipline four years of work experience in any sort of practical situations where you were able to address business questions and apply statistical analysis and you have some leadership attributes and you have solid communication skills so this is a general qualification of a data science position at google now let's dive into the interview process itself there are total eight steps in the interview process and it usually takes about six to eight weeks to complete the first step is a resume review followed by a chat with the recruiter pre-screened questionnaire technical phone screen on-site interview hiring committee team matching and the last step is probably the most exciting part which is the offer packet now let's walk through each step of the interview process in details the first step is the resume review so this is the part where once you submit your application on careers.google.com or job platforms such as linkedin a recruiter will review your resume and they'll see if your profile is a good fit with the job position that's available now they look for a couple things and they look at things such as your years of experience work experience academic background and your skill sets and if they seem like it's a good fit then they'll reach out to you the next step is to chatting with the recruiter so this is usually a 15 to 20 minute phone call with the recruiter and you should expect to discuss following information such as your background career interest and profile fit and then there are other topics that you could potentially discuss which include company culture benefits location and the hiring timeline now when you talk to a recruiter make sure you provide a compelling story about yourself and what your career aspiration is why you want to work at google make sure you also project positive energy because recruiters are also looking for your googliness meaning are you the type of person who is going to be bright and open-minded among colleagues at google and lastly use this phone call to understand the interview process as much as you can know what kind of interviews you have lined up along with information about what type of behavioral and technical topics will be covered during the interview the more you know this the better you're going to be in your preparation for data scientist interviews the next step is a pre-screen questionnaire so after you have spoken with the recruiter they'll send you a questionnaire for you to address so these are questions that the hiring team the recruiter hiring managers and hiring committee used to evaluate a candidate during the interview process and they'll include questions covering location preference timeline the tools and techniques that you possess and all of these information will be part of the submission to the hiring committee in terms of whether they should hire you or not within google once you submit the pre-screen questionnaire the next step is the phone screen so this is a technical round interviewed by a google data scientist who's usually l4 and above meaning that they're mid to senior data scientists at google and the entire interview itself is 60 minutes and it usually covers couple topics so it can be on your project background so this might be 10 to 15 minutes discussing the projects you worked on based on what you mentioned in your resume and definitely expect some follow-up questions from the interviewer and you'll be asked technical questions where you have to be able to solve them off the cuff so these questions can cover applied statistics data intuition and computational skills so all of these four topics could be covered within the first technical interview the next step is the on-site round this is probably the hardest part of the interview process itself because it consists of four to five technical rounds plus a behavior round and the interviewers are conducted by google data scientists who are l4 or above levels and also managers and each round is usually 45 to 60 minutes and a simple on-site rounds can consist of the following statistical programming applied statistics data intuition behavioral and a case round where they will give you some sample data set and tables for you to answer an open-ended case problem and present a solution now that you have completed the hardest portions of the interview the next step just involves a waiting game the interview panel basically all of the interviewers who have interviewed you in the behavioral technical rounds are going to send a evaluation of basically the assessment a grading score along with the recommendation on whether you are a strong hire or weakar to a separate committee which is the hiring committee and the outcome from this committee will basically decide whether you are recommended to be hired or not but it is important to note that this is not an offer just yet there is a next step which involves finding a team based on the skill set and the interest that you have provided to the recruiter the recruiter will find potential teams that you can match with so you will get a chance to talk to the hiring managers to see whether you would be interested in a particular role and once there's a fit between what you're looking for and what the hiring manager is looking for from there in the last step you'll be extended an offer and this offer packet contains information about the role the level the location and total compensation or tc and this tc is broken down into couple things so your base salary your bonus plus a sign-on bonus if this is something that you have negotiated and then equity as well which usually bests in quarterly let's jump to the next section in the video which is about the areas that are covered in the behavioral technical interviews so there are five areas that are covered ranging from applied statistics data intuition computational skills communication and googliness and leadership let's go through each area at a time the first area i'd like to discuss in details is the applied statistics the applied statistics usually cover the following basic topics probability hypothesis testing modeling experimentation and regression keep in mind that the interviewer doesn't expect you know advanced topics such as sequential testing or generalized linear models unless those are details you have added into your resume rather what their focus on is can you take these basic concepts and be able to apply these in practical settings so here's what an example problem can look like how would you explain 95 percent confidence interval to a non-technical stakeholder the next area i'd like to cover is a data intuition this is a part where the interviewer wants to understand your intuition and application of data and what that means is that they want to see if you were given a problem with some raw data set that's often somewhat messy whether it has missing value or it's highly skewed how would you use this data as a way to address a specific case problem in the form of statistical analysis or statistical modeling so an example problem might look like the following suppose that you have three continuous variables x1 x2 and x3 and you want to estimate the mean of x1 but 30 of the values are missing how would you solve this problem now toward the end of this video we're going to go through the solution so definitely stay tuned and the next area i'd like to cover are computational skills so this is another technical competency covered in the google data scientist interview and usually there are two types of questions so the first question could be statistical coding in the form of python or r and this is the part where you might be asked to do some simulation exercise and you might be asked to write a code that performs a specific statistical function such as performing the correlation coefficient and then there's the sql problems which contain tables and then you'll be asked to solve some exercise problems an example question covered in the computational skills round could be the following simulated code that shows that when the confidence interval is 95 about 95 percent of times the intervals contain the true population parameter so in this type of problem you'll be asked to solve it using r or python coding the next area covered is communication so in google's interview there is no dedicated communication interview rather in communication is an area that is covered across all of the interviews from the initial technical phone screen all the way to on-site in both the technical and behavioral realms every interviewers will be assessing your communication skills in terms of your focus clarity and structure of your responses so as you go through the interviews and you're providing responses to the interviewer's question you think about these qualities as you're addressing the questions the last area covered are your googliness and leadership so typically this is measured in the behavior round of the google's on-site interview and the qualities they are assessing are your decision-making leadership teamwork and attitude an example problem might look like the following tell me about a time when you pitched an idea and how you convinced your team to use it what was the impact so you want to think about a particular story based on your career experience or academic experience and provide a compelling story that addresses this question now as you're preparing for google's data scientist interviews i like to provide tips that can be really helpful for you the first tip i like to provide are you want to definitely brush up on the fundamentals so it's not that you're asked some advanced questions such as you know what is generalized in your model or what is markup chain unless you actually put them in your resume then you might be asked these type of interview questions but for the most part any fundamentals in statistics and machine learning and coding often covered in undergraduate level courses will do just fine in terms of being able to address most of these google's data science scientific questions the second thing is practice mock interviews so when you go on technical interviews you're going to feel a little bit nervous and at the same time you might not be ready to be able to address questions off the cuff and so that is why it's important to practice mock interview by pairing up with a study buddy through college or a friend that you know or you can even sign up for datantv.com and you'll be able to find a study buddy through the monthly subscription course the third thing is you know be positive and excited these are some of the qualities that the interviewer is looking for whenever they're measuring googliness they want to gauge whether you are excited about the opportunity and if you come across pessimistic then it's not it doesn't provide a good experience for the interviewer but at the same time it's a little bit of a red flag in terms of whether you're a good fit in google's culture tip number four is to ask clarifying questions so more often than not candidates will dive into a solution or they'll throw some ideas out there without fully understanding the problem asked a lot of these questions asked in google's data scientist interviews are not definition based questions like what is p-value there are often some business case problems where you have to be able to first of all clearly scope out what is asked and what is an answer in that problem statement and then from there provide a solution that is why it is important to ask clarifying questions and that is something that in the interpreters of the board tip number five is to provide a structured response so when you're providing a solution to a lot of these case problems it's really easy to ramble and so that's why it's important to once you understand the question that is asked you in the form of clarifying questions from there take a pause and then think about what is a structure in terms of addressing this question and the more structured response the more the interviewer can see that you have expertise in the domain that they are asking you questions on tip number six is engage the interviewer so sometimes candidates will just ramble and they'll kind of go on this monologue without taking a boss to just pulse check with the interviewer don't do that just make sure you interact with the interviewer if you get stuck ask them some suggestions or guidance and then from there continue with the response keep in mind that interviews are not just questions and answers they're essentially conversation the interviewer wants to get a glimpse of what it would be like to work with you as a colleague at google and so providing that experience by engaging them in a conversation rather than a monologue and the last tip is have fun i know that for many of you out there you aspire to work at google you know maybe some of you like myself had dreamed of working at google as early as high school 10 years ago but it's important to know that at the end of the day it is not be-all end-all if you don't make it and there's always another chance even if you fail google's interview the first time around you will get another shot the next year and the following year and again and again now i have a friend who has bought a work at google for many many years and he finally had a chance to interview for them and he studied two months straight nonstop only to fail the interview in the first phone screen and he was devastated but a couple years later he was given another shot and he became a data scientist at google so don't sweat it too much about failing this particular interview if you fail it you're always given another chance now in my video i'd like to share details that are comprehensive and invaluable so that you are well prepared for your data scientist interviews so in the last section of this video we're going to go through a sample interview question based on google's data scientist interview and the question is the following suppose you have three continuous variables x1 x2 and x3 you want to estimate x1 but 30 of the values are missing what would you do in order to address this question you want to start by comprehending the problem so there are two specific details from the problem statement you need to focus on the first is that you have three variables x1 x2 and x3 so something about x2 and x3 you can use in order to address this problem and then the second detail is that 30 percent of the x1 values are missing so based on these key details you can think about two important things so the first thing is that you want to consider the consequence of taking the average on x1 with 30 of the values missing that's quite a lot so if it was perhaps 2 or maybe at most 10 percent then taking the mean of x1 might be okay but 30 percent missing is quite substantial and so just taking the mean of this might result in some bias estimation what x1 could be so in order to improve the estimation of the mean of x1 what could you possibly do what you can potentially do is an imputation technique using a regression model where you're predicting what x1 is given x2 and x3 variables so you can train a regression model on the known values of x1 and then infer what the missing values of x1 are given the values from x2 and x3 and lastly once you have populated the missing values of x1 the next step is to essentially take the mean of this x1 with the imputation values so there you have it guys this video covers the end-to-end process of how to crack the google's data scientists interviews in the upcoming videos i'll cover more information about how do you prepare for product analysts and machine learning engineer interviews at google for more prep content like this definitely check out data tv.com where it contains courses and coaching services that are helpful for your interview prep i'll see you in the next video bye

Channel: DataInterview

Views: 1,395

Keywords:

Id: 3qcnfThYNcg

Length: 19min 25sec (1165 seconds)

Published: Thu Nov 04 2021