Top 10 Myths About Finding a Job in Data Science — Part 1/2


so in this call way we’re gonna do it is
I’m gonna go over ten minutes about finding a job as data scientist
essentially tell you why they’re wrong and then also what you can do instead awesome so if this is your first time
welcome to the mentoring program super excited to have you here in this program
you know I talked about how to get a job as a data scientist how to be successful
in interviews also how to be successful in your career and a lot of other cool
stuff so it’s really about finding out where you’re struggling how I can help
you and giving you advice to overcome that and really helping as many people
as possible so these are super fun for me and we’re gonna go through today ten
different things that people think about data science interviews they just really
aren’t true and then what you can do instead to be successful let’s get
started the first one here is all the time from aspiring data scientists they
say I’m not getting hired because I don’t have enough experience this is
just an excuse trust me like everyone starts with no experience
right so everyone before you has started with no experience and they’ve somehow
made it this can’t be the reason you’re not getting hired it’s not because you
don’t have experience or you don’t have enough experience the real reason is
because you’re not able to demonstrate the value that you can bring to a
company that’s the real reason so there’s really two things you have to do
to overcome this this myth of no experience first thing you have to do is
get noticed so of course it’s always gonna be a challenge to demonstrate your
value if you can’t get anyone to talk to you if you don’t get anyone to notice
you so the first thing you have to do is actually get noticed right you’ve got to
you’ve got to reach out to hiring managers you’ve got to reach out to
recruiters directly message them tell them that you’re interested in the
company right have a nicely designed resume that’s gonna stand out because
trust me when data science jobs open up there’s gonna take two or three hundred
applicants everyone’s resume looks exactly the same so the first thing you
have to do is get noticed all right second thing you have to do you have to
demonstrate your value to a company no one cares if you have experience like
here’s the secret literally no
nobody cares if you have experience if you don’t have experience it doesn’t
make any difference and I’ve been I’ve been charged of multiple departments
multiple teams I hired 30-plus data scientists never
did I think like man I really need someone who has a lot of experience like
why why would you need that the reason people think that they need experience
is because they think that it might be a good indicator that that person is gonna
be successful helping drive value at their company all people care about all
companies care about is making money right that’s all it’s all they care
about you need to convince them demonstrate your value that you can help
the company make more money or reduce their costs and then from a human
perspective there’s a person that’s going to hire you and they want to hire
you because not only are you gonna make the company successful you’ll make them
look good the boss like the easiest thing you can do to be successful in
your careers make your boss look good it’s really the easiest thing everyone
starts with no experience it’s an even playing field all right you’ve got to
prove you’ve got to prove that you can break in okay all right number two that
you have to master all the technical skills let me tell you why this one’s a
myth because there’s way too many technical skills for you to ever
possibly master them before you become a data scientist a lot of people become
overwhelmed they’re like I got to spend a year learning linear algebra and
calculus and probability and statistics you know maybe scientific computing and
then I’ve got to learn you know inferential stats I learn Bayesian stats
I got to learn machine learning I got to learn the reinforcement learning deep
learning all these different things I got to be able to clean data I got to
learn you know cloud technology all this stuff and they get overwhelmed to think
that they need to learn all of it and that’s simply not true there’s just too
much to learn no one’s ever gonna know it all especially if you’re starting out
people don’t expect you to know everything instead of trying to master
everything just get good at one thing I know I’m writing sideways here but you
really just have to get good at one the reason you’ve got to get good at one
thing is you’ve got to have a way to differentiate yourself from other
candidates and a way to show that you’re gonna uniquely bring value to a company
so maybe it’s getting good at NLP maybe it’s data cleansing maybe it’s unit
testing maybe it’s building data pipelines
setting up cloud technologies could be deep learning but it doesn’t have to be
like there’s a lot of different things that you could do it could be domain
knowledge could be they’re good at framing up business problems alright
were understanding business processes or maybe you know a lot about the insurance
industry it really can be anything but you’ve just got to have one thing that’s
gonna help you stand out right like that’s the critical thing and just know
know the basics so you can get through some of the basic interview questions
because you’re gonna learn most of the stuff on the job anyway
like that’s that’s the biggest reason that this is a myth is because you’re
going to learn on the job cool so let’s move on to number three
we’re gonna have to speed up here deep learning is an absolute necessary you
need to know deep learning to get a job in data science and this is just silly
most companies have absolutely no use for deep learning at least not right now
most companies you can think of them as like infants and toddlers in terms of
data science trust me every almost every single company is totally over selling
themselves and their capabilities and most companies and industry are way more
basic than you think they are like if you saw what I BM is actually doing it’s
not nearly as cool as they make it seem on the commercials most companies like a
linear regression would make them a lot of money properly gathering data doing a
simple regression doing some simple predictions would be worth a lot of
money to a lot of companies and there’s really no use for deep learning with
most companies yes of course there are exceptions like you’ve got way mo right
so that’s Google’s self-driving vehicle division yeah of course it’s gonna be
important for them but it’s such a small percentage of the marketplace it’s just
not a necessity if you’re starting out don’t don’t spend your time and effort
trying to master deep learning one like it’s not really necessary for most
companies there’s not really in a use case there’s not a lot of unstructured
data they’re trying to solve problems with like they’re gonna start with data
that’s in their warehouse already that they’ve already got stored in sequel
probably using Oracle or some legacy solution it’s going to be tabular data
so you’re not really going to have any use case for deep learning and second
it’s gonna be really tough to get I in from executives and stakeholders if
your model is not clearly explainable like what’s actually happening in it
people like to understand something so that they have confidence in it like
people have more confidence when they know how something works and then
they’re more likely to greenlight it to give you budget to push it into
production you know to let customers be exposed to your predictions so don’t
start out with deep learning not really necessary and if you talk about it
you’re fine because you say have to pass some screening questions you’re not
gonna go really deep on this stuff for most companies perfect so let’s let’s
keep rolling here to go and like a follow up with this one is you see this
everywhere and it’s kind of funny because you’ve seen it everywhere for
like the last two years you’ve gotta know Big Data tech you’ve got to
understand gotta be able to work with Hadoop and hive and spark and all this
big data tech and again most companies the huge majority of the marketplace way
more basic right they’re starting from the ground up still they’re really
struggling to drive value from data science teams and the two biggest
problems are one gathering data to deploying a model it’s not dealing with
big data trust me that’s not the biggest issue a lot of times you just don’t need
distributed computing because you can get a hundred get you know a system with
100 gigabytes of RAM or like a thousand gigabytes of RAM and so many cores that
you just don’t even need it for most problems and you can sample your data
set and you can segment and you can build multiple models you can do all
these different things a lot of companies just have a whole lot of
medium data right there’s a huge difference between having a big data
problem and having big data that’s just a lot of medium to small problems
because an aggregate it doesn’t matter you just have separate you know separate
CPU based machines to work on these problems and it’s fine so most companies
don’t really need big data technology so again literally just do the minimum to
check the box so you can talk about this in a job interview let’s move on here
alright so here’s a good one number five resume must be 80s friendly like people
are obsessed with this it is so weird to me so if you’re not familiar ATS
applicant tracking system so this is the system these are the crappy online
systems where you go and you send your resume into a black hole and then you
get an automatic rejection a few days later these are the applicant tracking
systems so let me give you some inside information as a hiring manager nobody
gives a shit about these be like if you think that these are stupid as a
candidate trust me every hiring manager thinks they’re
stupid too they don’t care about them I don’t think that I’ve ever sourced a
candidate that I hired from the applicant tracking system it’s basically
pointless these things just automatically reject everyone there’s no
like there’s no correlation that candidates that get accepted by the
system versus rejected I don’t see any difference in the quality of candidates
so what I would do is I would just go and download the entire list of resumes
and look at all of them I would just totally ignore whatever scoring weird
thing this thing does I don’t care I never use it
and then I’ll just directly contact the candidate because this is basically just
a big waste of time anyway yes you need to apply for jobs because that’s going
to be a rule at a lot of companies like we can’t hire an applicant if they don’t
go through the application process so you need to apply for jobs but then you
need to contact you need to contact the hiring manager you need to contact data
scientists on the team you need to contact HR you need to contact everyone
at the company and let them know that you applied for the job you’re
interested in their company you’d be a great fit and you think that you could
really help bring a ton of value to them right and then they’re gonna dig up your
resume from this ATS they’re gonna pull it out and they’re gonna click the
button to move forward with an interview so don’t even bother trying to optimize
for this this is something that I tried to do a lot as a candidate when I was
starting and it’s literally just a massive waste
of your time because I saw as a hiring manager these things are so clunky no
one wants to use them they never work they’re all junk so don’t even bother
trying to optimize for it because you’ll just never win you’ll just waste your
time the genius tailor your resume for every
job you need to tailor your resume and your cover letter for every single job
for the best chance big waste of your time what I see a lot of candidates
doing as a hiring manager what they like to do is they get the job description
and I say well look the manager is looking for a B C D and E so then they
go and get their resume out and they write down ABCD and E and they say look
I’m the perfect fit and me as the hiring manager I’m like oh great I’ve got
someone who just copied my job description
and put it on the resume they must think I’m an idiot like trust me people know
when you do that people know when you’re modifying when you’re gaming your resume
to look like the Job Description hiring managers are smart they will figure out
what you’re doing in less than two seconds so don’t bother trying to make
your resume look perfect like that make yourself look perfect for that role put
your best experience on there put your great projects on your resume because
those are gonna be more valuable that’s what you’re gonna talk about more so one
thing you can do is just kind of shuffle the order a little bit like if you’ve
got if you got two projects here I call this one P 1 and you have P 2 if you
think that one of them is like much more relevant for a job you can maybe switch
the order of them so that whoever’s reading this will see the other project
first because it’s more relevant for them but don’t go changing the language
in your resume putting new projects on it and having like 20 different versions
of your resume this is a huge waste of time it’s pretty much always gonna
backfire everyone’s gonna figure it out you’re not clever people know what
you’re doing trust me they know so just put your best
foot forward put your best projects on there and just use that resume for every
job and then don’t even bother with a cover letter cuz no one reads them why
because I get 300 resumes do I want to read 300 resumes and 300 cover louder
like no I’ve got a job to do I can’t spend all week just reading cover
letters so no one bothers no one reads them they don’t matter I never wrote one
even when they were required I’d write like two sentences I’d like copy my
LinkedIn profile and be like there’s my cover letter data scientist I like data
science like it can be so trivial it doesn’t matter because no one’s gonna
look at your cover letter so don’t try to game your resume you’re not fooling
anyone and don’t bother with a cover letter it’s not gonna help you just have
one resume that’s a really good that shows off your best skills and apply for
jobs that you think are reasonably appropriate for your skill set all right
so we got a lot of stuff here make sure that you’re applying it and don’t get
frustrated if you’re stuck stay consistent keep working through it
everyone gets frustrated everyone gets stuck the people that are successful are
the ones that keep pushing through and don’t quit that’s really what it comes
down to keep learning and proving don’t quit you’re gonna do great and I can’t
wait to see you on next week’s call

Leave a Reply

Leave a Reply

Your email address will not be published. Required fields are marked *