Interview: Amazon Business Intelligence Engineer (Business Intelligence vs Data Science?)

i'm a time saver i am a magician i'm a tester i'm a database administrator i'm a marketing person i'm a sales person bashash can you tell me a little bit about yourself i'm honored to be here young i appreciate you thinking about me so it's really cool of you uh my name is vashesh shankar i'm a business intelligence engineer at amazon i am rounding up my first year here i've been in the bi space for this will be my ninth year actually this september so i've been in the bi space for i think before it was called bi and before it was really about analysts and data engineering and so yeah i've been there during inception of bi like what are the roots of someone getting into this field like is there a specific major in the university or college or do you just one day say like hey like i'm gonna be a bi so this is one of my favorite questions to ask answer for folks when they ask me this like hey how'd you get started what's the traditional route let's just uh bucket it as my route was not traditional at all i actually was a entry-level job at nordstrom i was walking around the hall one day and my my manager soon to be a manager walked up to me he's like hey do you have a moment to talk and i'm like i'm using the bathroom what's going on so i need to talk to you so we sat down for an hour right there uh i vividly remember the conversation he goes hey do you know anything about sql i'm like what's that do you know anything about uh excel you're like barely what do you know about tableau nothing because the tableau was brand new at that time um and he's like look i heard you do good work if you come work with me we'll teach you skills to learn sql we'll teach you skills learn vba which is an object-oriented programming language for excel from microsoft office um and he's like we'll teach you these skills in about three to four years you'll be making sixty seventy thousand eighty thousand dollars and i was like i could pay more money to do the same same job eight to five great sign me up so that was the beginning of my journey um and i just worked my way up through there i think it's important to preface that with i have a degree in mathematics and economics so i definitely have a leg up and folks in terms of just like math and and business acumen knowledge um but i don't think that is a prerequisite for this field i i have one of my my favorite mentors i've had he has a degree in psychology he's actually here at amazon as well so uh there is no set path what i think is important for folks and i always tell us to people i don't care if you're going to bi if you're going into vendor management if you're going you could be going into doctor being a doctor i think it's always important for folks to understand a little bit about code uh whether that's sql which is a a declarative programming language to help describe the data that you want to return or you look at python or c or there's so many declarative programming language where you're explicitly telling the computer what you want the computer to do knowing the baseline of that is going to be honestly an entry point into business intelligence because business intelligence is a it's a field where we kind of hit on a multitude of subjects we're hitting on data and analytics so we need to understand the data we're hitting on data engineering understanding how the data needs to be shaped and sculpt data science sometimes i'm doing predictive stuff i'm trying to predict what my customer is going to be looking for looking after and there's also the business intelligence it's kind of melding that together speaking the business side speaking the tech side getting the two to work in kahoots and then oftentimes translating one or the other to the other side is the challenge there so that entry point is going to be very broad but once you get into it i think things that people should be looking out for like oh i i think i would be interested in business intelligence if you have attention to detail if you like to find patterns if you like analyzing data and deriving insights from those those are more of the skills you want to have they're going to be applicable to other realms as well you can be a program manager you can but definitely if you like getting into the nitty gritty of the data understanding how you know a shirt sold on one day and how that is in relationship to how this the shirt category is selling for the entire year and those type of relationships definitely going to be a good starting point for me there but again i preface that i'm not the typical come from school and make your way into bi but i think that's important to highlight there is no one path that's all for bis definitely like build your own path understand where where you have your niches and what you're interested in but if you're interested in data bi and you're interested in business bi is definitely that perfect realm i interact with both technology and interact both with the business and it's a good amount of both and i have to be able to speak in both languages very deeply and very fluently gotcha you i think you kind of mentioned it while you were answering this question but i also want to learn a little bit more on like so what's the difference between someone who's in bi like business intelligence is like someone who's in data science yes yes so actually i spent a lot of time researching this uh last year because i was trying to get a business intelligence engineer title i thought it's very important and in that understanding i came to a the best way that i can answer that to me if we look at a spectrum there's data analysts there's business intelligence engineers and there's uh data engineers and data scientists those are kind of the realms you're going to be looking at and they kind of sit on a spectrum and really business intelligence sit underneath all three of those and let me explain the three above so data analysts are people that are telling you why something happened they will look at the data they'll look at the reports they'll look at the facts and say hey because this product sold this product didn't or hey we had an impact in shipping availability so this product installed they're telling you why something happened they're reviewing past facts data scientists are trying to predict the future they're trying to say based on the models we set up based on what we've understand with our customers we're going to put it through this algorithm and this algorithm is going to say this item is going to be sold what's going to be sold at this point in time a data engineer is sitting and saying what is the fact what sold at what time at what point in time et cetera et cetera they're making that more accessible business intelligence engineers sit across kind of all three realms there i have to play a little bit of each one of those hats now personally i like the data engineering part the best because it allows me to really impact what predictive stuff we can do and what questions we can answer and that's where i feel like i bring the most value at least in my role but bies are a very broad category here at amazon they were just reclassifying the technology uh family and there's a lot of different entryways and a lot of different expectations with entering into the bi spaces what are your roles so what i've noticed here at amazon there's really three different types of bis there's going to be again on a spectrum there's going to be bis like myself i'm on a bi tech team so i have a bi manager so he is a former bie i sit on a team with six other bies and we do decentralized reporting needs for our organization the hard lines organization so we sit in a group and we're handling very specialized uh reports that are going to go out to the masses then on the other end of the spectrum you have uh someone who is uh a bie who's sitting on a retail team so they might be the lone bie working on a with a retail team and their functions might be more focused towards the data analysis side of it so they're going to be getting some data inputs and helping the business understand what should be the next step and that's where the but they still have to do some of the same category of stuff that i mentioned before and in between that there's going to be a bi on a tech focused team but they're the only bi on their team so that's a your your manager has some understanding of technologies but they're they're not focused so much on retail you're maybe doing bi in terms of how a program is working and it's there's that's where the line gets muddy because you're not working so much in business intelligence more maybe data intelligence but the term still holds true so again there's a bi is handling multiple roles and it really depends on what team you end up falling in which end of the spectrum you're going to be leaning on and doing more of those activities so i'll just ask you about your process of getting into amazon so i would assume that interview process for different roles for example for me marketing manager and ubi engineer would be very different can you explain a little bit did you have any technical assessments for example while you were interviewing yeah i would have no doubt in mind that if if our interviews are anywhere near a like someone's not doing their rock right but yeah so that's a great question i've actually also recently got more involved with interviewing so i can talk about my experience and then more how things are shifting as well so in for my interview i had a five-part interview loop uh three of those were non-technical and two of those were technical i think that's again very one of the unique parts about bi there's a good blend of we need technical acumen we need business acumen so there's going to be a lot of sussing out from those conversations and especially before what business impact have you had what technologies have you operated with so that's definitely the difference there i don't remember the particulars about my individual segments of interviews but i can tell you um just because i had non-technical interviews with three of those folks two of those folks were still asking how my technical job impacted my ability to do non-technical stuff so there's kind of that blend there so that's definitely something to be aware about as a bi we're going to be interested in how not only how you interacted with technologies and the reports and the data you created but how do you interact with your customers how do you work you know work with your customers or work against shifting deadlines as many of us know in retail the things change daily so we how do you operate in that space i feel like i don't know personally but i feel like at least in a tech space there's at least a bit more rigor to it and there's a bit more of a timeline and bi i could i could have a product launch today i'm gonna probably launch tomorrow there's no uh cadence there's no set cadence so it's a bit more uh flexible to what retail needs and so how do you navigate in that space those are some of the questions we're gonna be really looking for um in order to make sure that you're gonna be able to adapt to this fast-moving environment that we have here uh one thing that i would definitely say more than i'm seeing in in the interviews i'm now actively coordinating is we're doing more of a technical assessment and a bit more of a rigorous technical assessment that also leads way to the bi job family moving in technology i can only assume that job uh that that that assessment will only continue to evolve and get more robust as we as we fully integrate into the job tech the tech job family yeah shift think yours a little bit tell me what you do as a vi engineer i'd say in a normal day today there's a couple of things at least on my team uh the two things that happen weekly is we i host an on call so i will manif man of death so we're in virtual corona time so it's virtual but when we were in the office i'd mana dash for four hours any one of my retail teams and my my partner organizations like yours young have the opportunity to come to sit with us for uh we have a four-hour window where people come ask this question sql-based questions requirements face questions um and so that's four hours of the week we have another uh two sessions each an hour where the entire team shows up and we also assist with any questions across the entire organization so that's six hours of the week out of the four eight hours um there's another probably it's a four hours or just in in meetings and touch bases and updating status so let's talk about the other 30 and those other 30 um you're looking at a combination of things depending on where i am in the development cycle so if i'm starting a new project maybe i'm gonna be spending more time writing raw sql validating data points um making sure i'm building the right metric definitions as i get further along in the process maybe i've built the the etl extract transform load jobs and now i'm ready to build a dashboard i might spend more time working on as simple as stuff is building a simple dashboard formatting you know building making sure the experience works correctly there's time building documentation so a lot of the etl that i work on is complicated it's not something you can just read one or two lines that's a thousand lines long so documentation highlighting um pitfalls assumptions data metric definitions that are readable for a business user as well as tech users so there's that kind of like creating that that doc again we talk on this a lot how do i bridge that gap so i'm creating documentation that's usable for both uh both tech and business users i am a level five on my team and so as a level five i'm involved in doing the code review process so members of my team below and above me will will submit code and i'm par i then become latched on to one of those projects and i do a review on that code make sure the sequel's right call out you know logic issues et cetera et cetera like i mentioned to you there's there's a bi we wear a lot of hats it just kind of depends on where we are but again we're going to be one of those three am i doing some type of analysis am i building some type of environment to host the data or i'm helping build some type of future state so a lot of that data sciency because we don't do data science yet we're not building models yet on my team but we will do things to predict so the customer will say hey i'm going to take this report i'm going to do x y and z things and so i'm trying to predict those things oh you're going to take a support what are the next six steps you're going to do how many of those can we cut off your plate and put this as a bi process rather than a manual process that's where when i like to put a pause like that's what i officially do what do i like to say unofficially i'm a time saver so i'm often gone to to find ways how do i get data better how do i save time how do i process um i am a magician so i'll make things appear where they shouldn't or make things you know accessible places they haven't been before um i'm a tester so i'll be testing products i'll be testing our internal products i'll be testing our flow i'm a database administrator uh not so much in amazon that that role has been a bit split out but it could become back something that that was on my previous team i was database administrator for sure so i don't wanna that could be something you are responsible for here at amazon just depends on your team dynamics um i'm a marketing person i'm a sales person i have to sell my product if i create this awesome dashboard i gotta make sure people know i gotta work with the right people i gotta i gotta um evangelize it i gotta i gotta get it out there so sometimes it could be a big launch that's going globally or enterprise wide and it's easy because i have a team of people behind me helping me or sometimes it's just me i have launched it so i have to then figure out the best way to to get it out there so there's a lot of hats that we wear but again the key things are developing data sets developing uh dashboards that then our end users will be using that's like my you know top level but once you start to break down the cookie and get the crumbs that's what we're talking about yeah what would a bi's career projector look like uh i think the first thing always is going to be i think the business intelligence engineer role is a great way to understand what is my next step right you can you you dabble in so many small technological pieces and technology pieces as well as business pieces you kind of if you do join as a bi you do kind of open your realm you open your world like you could go down the product manager role you can go down a project manager role you can go down bi and advance in bi you can go down towards data science you can go towards data engineering there are a lot of avenues and so i don't want to say this is when i started you know this my initial intent intent was to become and eventually become a data engineer i joined and i was like i actually like being a bie i might want to stay as a bie so you have that flexibility i don't i feel like you're not pigeonholed as much because you can go to your manager and say hey you know i came to you and told you i'm really interested in de stuff but i'm now really liking this data science stuff and you can figure out with your manager how to get projects and things that can push you in that direction so i i don't think there's a typical b that's the beauty of bi it's very open-ended but i think if you the the general consensus if i were to you know screw it all down is you'd be a bi and eventually move into data science because they they tend to bleed and doing the same things and the most successful data sciences that i've worked with have had a baseline knowledge of what bis do and allowed them to get their data better and help build their models um but i only know one i've only known one i've known of maybe a handful data scientist so i can't speak to the data scientist community as a whole i just know the one bi data scientist i worked with he was able to get to the data he needed better he's able to build his models faster and he just had a better experience that's where i originally was thinking about going down that way but i love being a bi that's definitely a role i continue to i will continue to work towards what surprised you the most after joining amazon the most the most interesting thing that uh that came to me when i joined amazon the process of doing some bi related activities doing extract transform load that's a common activity for any bi a lot of that had been has been abstracted and so what i mean by that is it's taken out a lot of the heavy lifting i used to do my old job so one of the biggest things i had to do all the time was data quality checks the data properly come from the source systems and land in my local repositories and as my data quality good a lot of that stuff has been abstracted and there's a there's a web ui that amazon has set up that handles most of that i can offload that thought process and mind power to that system and so now instead of me getting being on call as early as six o'clock in the morning and wondering if data qualities if things ran correctly and all these steps that's no longer so i can spend more time is what i tell my other bis and my team i'm spending more time doing my job where it's most i get to use my brain and i get to focus on using my brain power instead of using my brain effort towards making sure things that should be in place which is data quality and stuff like that is actually accurate so it's just a way that was the best and the most surprising thing is i'm able to spend more time doing my job things that have value and adding value than me than spending time just doing maintenance activities gotcha yeah did was there any learning curve once you join um as i assume that we would probably have other systems or tools that other companies that does not use yeah i mean that's a that that my previous point bleeds directly into that there are certain things that if you know sql when you come here you'll be like oh that's sql great i can i can look at that and see there is definitely a learning curve learning some of our internal systems especially that web ui that i just called out um i think the other big thing is there's just so much data it's i mean i thought i i came from nordstrom before and i remember talking about how much data i played with and i'm sure my interviewers were just laughing like that's like a week for us buddy you know like just the volume is overwhelming uh and so that's definitely and then not only is the volume overwhelming but what's different also is how many people are creating data and so like the combination of the tube kind of creates an uh information overload like where do i go to get the right data where you know you obviously figure that out as you spend more time here i've been here for about 11 months so you know and i feel like i have a good understanding of what that process looks like now i think it took me about three months to really feel at home with that yes there's a learning curve to that uh the other thing i want to hit on is i also joined a bi tech team so i was surrounded by six seven other people and my manager especially who have been around the block that they know were a lot of those important tables important data sources so that also i think speaks to more why it took me three months to get ramped up versus i think a traditional six months in a bi space so that being said that that definitely is the just over like whoa but once you get a handle on it definitely you know it goes back to being sequel and you know you're doing your job so what do you like most and what do you like these about your girl i really like the uh for those of us amazon i might think it's cliche i really like the level of ownership that i'm i'm given and and my ability to take ownership of certain projects i think i i mentioned to you earlier in this conversation i've been working on redoing this entire dashboard that i was given um no one told me to do that no one asked i know they just said bring it over and make it part of our environment as like i i don't think this is something i want to take ownership of because it's not something that provides the right value and the right information right now it needs to be and so i went to my manager i explained to him situation i gave him a list like here's here's where i see those problems and he gave me the opportunity to actually make improvements he saw the value in that and that just just coming to him saying i see value and taking me taking true ownership and fixing this was this you know that's something new to me that i i can tell you multiple times in my previous role i'm like hey something's on fire over here something is wrong something is going to provide false false positive data et cetera et cetera uh and only for me as short as six months as much as i remember one situation it was four years later where someone's like hey you guys need to fix this i'm like here's my email from four years ago where i called this out and you guys didn't give me time to take ownership over this so like it's that type of ownership that is definitely like uh it's refreshing and and it's and people especially my managers trust me when i come to them with a problem or an issue that we need to take ownership of this we need to solve it it's going to get solved and we we you know massage the work around there this wasn't a project that my manager wanted to give me the time to do right originally once i articulated the value shared it with them we're able to then make some space and and get it along um things that i don't like i think the biggest thing and this isn't going to be here or at previous roles i think in general for some reason people think bi engineers are sitting around waiting for people to message them and say hey i have work for you to do so it's just one of those like i just because i i'm working on your project doesn't mean i have 20 other things on managing and so just because you know i haven't messaged you back or i think that tends to be um i don't think enough people know what we do i think that's my answer people don't know exactly what we do and so they just see the magic we put out there and sometimes we do things that they think isn't super complicated it takes us a day and they're like now everything can be done in a day versus i do this the thing that they think would be super easy and i come back and be like hey this is gonna be like three weeks um and so it's like understanding where and how long project and then the other problem and tricky part with my role is you will be given a task or an assignment people like all right how fast can get this done um and the thing about business intelligence is it's a perfect blend between the art and science right the business sides really are exciting the data is really the tech side really science side of it and so for especially for new bis giving that first estimate is like i don't know like i think it's going to take four weeks so you know and then you get done in a week and now your manager's looking like hey is your estimates going to be this off every single time or the inverse hey it doesn't take a week and now it's taken six weeks and so that estimation is such an art science um that was really hard when i came to amazon because i didn't know the data i didn't know the the the shape and what's going on and so um there's a lot of unknowns in business intelligence as you're starting to explore data and so that's just a common pitfall i don't think there's a perfect hey how do you how do you fix that just more experience um working with your with your partners who whether it's your managers your business partners to understand and try to get out as much information from them so asking a lot of questions is a way to help tackle that problem but there's never going to be a perfect answer gotcha my last question advice for those who want to join amazon or enter business intelligence yeah um let's break those down for when it comes to business intelligence again i'm gonna whether you're whether you're been in working for 40 years whether you're you know first year in high school i don't care what it is crack open some introductory coding classes that's first and foremost um sql and pick an object oriented programming language the hot one of course is python you can go java you can go do c sharp you can do c i mean pick one just pick pick sql and pick one of those programming languages and just mess around with it uh you'll what i think you will come to find is basically everywhere you interact with the web something more than likely is going to touch a sql database as they do a report on it if you start to get to that level of detail so just understand how technology is around you and your interaction with it and how the information you put on a web form might get transformed into data is a really important concept i don't care where you work because every whether you want to admit or not every company is a technology company these days so understand a little bit about how the technology the data you're creating and whatever you're doing whether it's a school project or a web form you're filling out or your job might get transported to a relational database somewhere understanding those relationships and then part b to that is understanding how code works understanding how a script might work so python or those things just the basics that's going to tell you okay if it interests you or you know that's going to tell you hey maybe i want to explore some more bi space if you open it and you hate the computer and all that stuff get away from bi it's not for you so uh and then it's getting into amazon i think the biggest thing when it comes to applying for amazon is i think it's best for folk this is the advice i give to all my people that apply is what you should do and it took me almost eight to ten hours to do this so it's not like this is easy but sit down think about the work that you've done in the last three to six years if your time if your experience has been shorter whatever then take each one of those projects and really think about did i own that project did i just maintain it did i was i just a tertiary did i just have knowledge of it kind of break it all out and then from there for the ones you owned you maintained you updated um whether it's a process put some facts to it i did this process in my old job by doing this process i saved x amount of time you want to come with these hard facts especially when you're applying to bi because you're around data you better have these facts so explain how what each one of these projects these processes do have that ready to go memorize it study it i don't whatever makes it so you have information because it it makes make sure that you know we know that you you understand the impact that you're having when you when you apply for that job it's not just like oh i made this process better okay what is better to use better five minutes is it an hour is it 10 hours talk us through that uh the other thing you really want to do is as as many of us know if you when you apply to amazon we have leadership principles here and definitely understand pick those projects you worked on especially for climber bi understand how those projects related to some of the leadership principles maybe where you can highlight where you showed a leadership principle in particular so like this project i just worked on i'm really showing ownership here and i'm showing that i didn't accept the previous version and i really want to make it better so i showed ownership and i'd almost say tag each one of these projects or these things you've worked on with one or two you know leadership principles just so you have a couple examples ready you don't have to say it in the in the interview or whatever but understand and kind of speak the same language as us and make sure that we are seeing it the same way you are and you can kind of put those tags to it um that's for me what i did and i you know going to my interviews it made me feel more confident that i was able to answer the questions at least you know talk about the the questions of the way the interviewers are expecting and have some ready to go examples in those genres and no some projects could touch on different genres so you don't want to end up talking about the same project for every single interviewer you have different ones and you're able to really adapt to each interview question yeah i think what you mentioned about um you know breaking those work you've done into what i owned what i improved and what i maintained like i think that's i've never heard about that but i think that really really helps because at the end of the day interviewer mostly asks you about like so like i got asked multiple times so what did you do you know like what did you do and what was the business impact of it and like given that you are you know interviewing with as a vi like that will be much much more important because you probably have to you know provide the data as well um but awesome yeah great great answers and thank you so much for your time today um the audience will be greatly appreciated about your information

Channel: Career School

Views: 24,481

Rating: 4.9590445 out of 5

Keywords: business intelligence engineer amazon, business intelligence engineer amazon interview questions, what is a business intelligence engineer, data scientist degree, data analyst qualifications, data analyst jobs, BI, Business Intelligence (Industry), Business Intelligence, performance management, business data, career school, business intelligence, business intelligence engineer amazon salary, business intelligence engineer amazon salary india, data scientist

Id: GeKyfcUn6JY

Length: 27min 13sec (1633 seconds)

Published: Tue Oct 13 2020