SDS PODCAST EPISODE 203 WITH SASHA PROKHOROVA€¦ · scientist Sasha Prokhorova. Kirill Eremenko:...
Transcript of SDS PODCAST EPISODE 203 WITH SASHA PROKHOROVA€¦ · scientist Sasha Prokhorova. Kirill Eremenko:...
Kirill Eremenko: This is episode number 203 with aspiring data
scientist Sasha Prokhorova.
Kirill Eremenko: Welcome to the Super Data Science podcast. My name
is Kirill Eremenko, data science coach and lifestyle
entrepreneur. Each week we bring you inspiring people
and ideas to help you build your successful career in
data science. Thanks for being here today, and now
let's make the complex simple.
Kirill Eremenko: Welcome back to the Super Data Science podcast.
Ladies and gentleman, I'm very excited to have you on
the show today. You can probably already feel the
energy that I have, and that is because I just literally
just now got off the phone with Sasha Prokhorova and
we had an amazing podcast session which you're
about to hear.
Kirill Eremenko: So, what did we talk about in this session? Well, first
off, what you need to know is that Sasha and I met at
Data Science GO 2018, which at the time when you're
listening to this podcast was just over a week ago. In
this session what I did was I asked Sasha about her
experience at the event. I found that this was a much
more interesting way to share with you some of the
highlights rather than me telling you the highlights
that happened at Data Science GO. It was very cool to
hear them from Sasha's perspective, from an
attendee's perspective. Through her lens you will see
what her takeaways were and what were some of the
key things that she learned from some of our speakers,
like Ben Taylor, Rico Meinl, Randy Lao, and some
other people. So, especially if you missed out on Data
Science Go 2018, then this will be a great opportunity
for you to catch up on some of the things, some of the
key takeaways that an attendee had from this
conference.
Kirill Eremenko: The other thing that we did is we talked about Sasha's
background experience and her journey into data
science. She's been learning data science for one and a
half years and she actually brought up the usual
concern that I hear that how do I get a job, it's very
hard to apply for jobs and get through and get
recognized and actually get invited and get job offers. I
challenged Sasha on that, you will hear it, I had a
whole rant on what I think on this topic and gave my
advice on this topic. So, you'll hear that and plus I
gave Sasha a challenge. A challenge to get her name
out there and skyrocket her career.
Kirill Eremenko: In this podcast, by the time you're listening to this, she
should have completed her challenge so stay tuned
and inside the podcast you will know how to verify if
she has or hasn't completed her challenge. I think
that'll be a fun game. And, plus, to make it even more
fun, during the podcast I announce the same
challenge but in which you can participate in and
there's a prize. There's a prize for the person that will
do the best job on this challenge and you'll learn all
about the details throughout this podcast and the
prize is something that you don't want to miss out on.
It's something that will help skyrocket your career and
take it to the next level.
Kirill Eremenko: There we go. That's what this podcast is in a nutshell.
I'll leave a cliffhanger like that for you and without
further ado, I'm going to introduce to you Sasha
Prokhorova, an aspiring data scientist.
Kirill Eremenko: Welcome, ladies and gentlemen to the Super Data
Science podcast. Very excited to have you on the show
today and we've got a very special guest, Sasha
Prokhorova, calling in from San Francisco.
Kirill Eremenko: Sasha, welcome to the show. How are you today?
Sasha Prokhorova: Doing wonderful, thank you very much, Kirill. I'm very
happy to be here.
Kirill Eremenko: Awesome! Very, very cool because for everybody out
there, we literally just met with Sasha five days ago at
the Data Science Go 2018 event and it was legendary.
Had such a great time. Sasha, tell us a bit about Data
Science GO 2018. How did you enjoy the conference?
Sasha Prokhorova: It was such an amazing event. I got to meet and
connect with a lot of interesting people in the industry.
It was not the usual format of the conference that I
was used to. [crosstalk 00:04:24] There was a lot of ...
(laughs) ... A lot of informal aspects, for instance, all
the speakers were so personable and approachable
and we started out the day with a little yoga and
meditation, as well as a little dance. I thought that was
refreshing.
Kirill Eremenko: That's awesome. That was very planned and also I
think it went very well. A lot of people were enjoying it
and opening up. Do you feel like you opened up? Do
you feel ... Did you actually feel the energy in the room
go up after all of those informal elements?
Sasha Prokhorova: I did, certainly. I felt very inspired.
Kirill Eremenko: I had attendee come up to me at the end of day one
and he said ... well, at the end of Saturday and he said
that "Hey, Kirill, the energy's so good here that, so
high that I only now realize that all I had for food was
a sandwich in the morning." And then he even skipped
lunch even though there was, like, there was a
network he liked, he skipped lunch because he was so
into talking somebody and then he realized that he's
not hungry and he's not tired simply because of the
energy in the room which I totally loved, really loved
everybody contributing. I think it was like a
community effort in that sense.
Sasha Prokhorova: That's actually how I felt during, pretty much, most of
the day out there at Data Science Go. I almost felt like
I could have forgotten to eat because I just so
absorbed in meeting people and talking to them and
learning new things. That's the fear of missing out in a
nutshell. [crosstalk 00:06:01]
Kirill Eremenko: Nice. Tell us what was your favorite talk?
Sasha Prokhorova: I really enjoyed the talk by Ben Taylor. The opening
phrase was the market does not give a crap about your
dreams. I thought it is very true because it's not about
people seeking the opportunities on the market. It's
what the market needs and this is what the market is
going to select. The market is going to select the people
that are right for solving certain problems [inaudible
00:06:38] particular company. It's all about the needs
of a particular company.
Kirill Eremenko: So what was your main takeaway for your career from
that phrase from Ben Taylor's talk? 'Cause it sounds
like maybe for somebody listening to the podcast who
wasn't at the conference, it sounds a bit like, I don't
know, like pessimistic that the market doesn't really
care about your dreams. I think Ben put it ... started
off like that but then he explained it in a way that
sheds light on the whole thing. What was your main
takeaway?
Sasha Prokhorova: Important to have a passion and even borderline
obsession. My main takeaway from this talk is the lack
of experience is not really the end of the world because
before when I was looking for jobs and internships in
the industry, I was getting a little frustrated in it
sometimes because you would need job experience to
acquire job experience. It's almost like needing a pair
of scissors to open a box that scissors came in.
Kirill Eremenko: That's a great analogy.
Sasha Prokhorova: It's almost like, given [inaudible 00:07:45] and that
circle and Ben Taylor's talk gave me a really good
insight about how to break out of the circle. For
instance, it said lack of experience is a crucial but if
you're capable of doing a project, my takeaway from it
is just find a data set that you're passionate about and
pick a data frame, decide what to do with it, and
showcase your work. Showcase your work to a
potential employer or to all those followers on Linkedin
or even showing the family. Just get your work out
there and show that you did something productive
with your time. That you learned, you dared, and you
achieved.
Kirill Eremenko: Fantastic. Love it. Tell us who else, who else, who's
else's talk did you enjoy? Because you were there for
the training sessions on Friday, but then the main
event is Saturday, Sunday. That's one and a half days.
I think we had close to 25 speaking sessions. Who else
did you like? Who else did you love there?
Sasha Prokhorova: I really enjoyed attending the talk by Rico Meinl. My
favorite quote from him is "What is possible is often
limited by how hard you try."
Kirill Eremenko: Wonderful. Rico, I heard he did a fantastic job. He flew
all the way from Germany. Do you know that Rico was
an attendee last year?
Sasha Prokhorova: No, he did not mention that.
Kirill Eremenko: So he was an attendee last year, DSGO 2017, and
then during the event he came up to me and he said,
"Kirill, I want to be on this stage next year and I want
to help inspire people." To that I said to him, "Hey,
Rico, that sounds really cool, but you need to prove
that you can do it. That you are going to actually bring
value to people." And so, what he did is he went back
to Germany. He started an AI meetup, which is now
attended by several dozen, if not a hundred, people,
it's just like once a month, once every several months.
So, a meetup on AI [inaudible 00:09:56] learning, then
he introduced artificial intelligence in the business
that he's working for and the company he's working
for.
Kirill Eremenko: He did quite a few cool things like presentations on AI
and things like that, and then he came on the podcast
and when he told all, all this, he told me all about this,
I was like, "Rico, you have to come to Data Science
GO. You have to present." He didn't take it lightly, that
invitation lightly. He actually prepared his talk and
then he hired an acting or like a speech, speaking
coach, who helped coach him how to do this talk. So,
this guy's really serious about the things he gets
started in and, hence, the result. Everybody was very
impressed with his talk.
Sasha Prokhorova: Wow! His dedication is truly admirable. He's such an
inspiration for all of us.
Kirill Eremenko: Yeah, he's wonderful. Wonderful. Ben Taylor. Rico
Meinl. How 'bout influencers? How 'bout people that
you got to meet there, like who are also giving talks,
but did you have ... Were you excited to meet the
people that you follow on Linkedin in person?
Sasha Prokhorova: Absolutely. One of them would be Randy Lao. He's a
great resource to follow for those people who are new
industry, in the industry of those aspiring data
scientists. He's posts are just so informative and
incredibly concise and it's basically just a how-to
instruction. The algorithms that you need to learn. The
books you need to read. Just very on point.
Kirill Eremenko: And what was he like in person? Was he different to
what you were expecting?
Sasha Prokhorova: He was very nice and approachable and kind. He was
very appreciative of all the attention.
Kirill Eremenko: He's a very, very cool guy and I ... What I've found
actually during the whole event was that most or all,
almost all of the influencers that, who were there from
Ben Taylor, Randy Lao, we had Nadieh Bremer, we had
lots of ... Terry Singh, all them were very humble. They
were very open to talking and giving advice and
connecting with people and hearing attendee's stories
and just getting into this community and really giving
back. So that's what I really appreciated from them
and I think it resonated well. There were so many, so
many great conversations. What was the most
surprising thing that you learned at the conference?
Sasha Prokhorova: The most surprising?
Kirill Eremenko: What was most impressive? Something that really got
you inspired and, apart from the talks, I mean during
networking opportunities with people?
Sasha Prokhorova: That everyone was really approachable. Data
scientists, data engineers, you have to remember that
they're people at the end of the day, very brilliant and
outstanding people but they're people and it's
important to connect. It's just important to get yourself
out there, no matter how shy you are, and no matter
how hard introducing yourself and talking to people is.
It's really important and actually just get yourself out
of the comfort zone.
Kirill Eremenko: That's a very good point. Oh, and one more thing I
wanted to ask you. As a woman, what did you feel
about how represented women were at the conference
in terms of speakers and in terms of attendees?
Sasha Prokhorova: The demographics were very balanced. There were a
lot of women who attended and I definitely felt a lot of
support from everyone in the industry regardless of
gender.
Kirill Eremenko: That's very, very good to hear because one of our, one
of the things that we're trying to improve and change
is the status quo. In data science, typically, it's about,
the ratio of male to female is about 90 to 10. So 10%
female in the industry, about so, but at our event for
instance in terms of speakers, we had 35% female
speakers and in terms of attendees, I don't have the
numbers yet but as soon as we have the stats, I'll
announce them as well. I think, I think we did quite
well in that sense and it's important to inspire and
show role models for aspiring data scientists. That,
regardless of your gender, race, background, you can
succeed in data science. That's, I think, is good to hear
that you felt that at the event.
Sasha Prokhorova: No, absolutely.
Kirill Eremenko: Well, shifting gears, thank you very much for the quick
overview of DSGO and your experience there. Let's now
move on to your journey through data science. So, one
of the reasons, for our listeners out there, one of the
reasons why I decided to invite Sasha to the podcast
was when we met at Data Science GO, I found her
story quite inspiring. Actually, very different to what's,
very unique, I'd say, or unique and quite inspiring for
many people out there who are starting into data
science or who are already in their journey in data
science and want to look back and see how it was to go
through it.
Kirill Eremenko: So, in short, Sasha will give us a background just now
but Sasha's in a bit of a different industry. She's now
specifically in data science. She's an electrical
engineering student, but Sasha feels the importance of
knowing data science and integrating it into her
career. So, that's what I want to dig into a bit further
and why you feel that way and how you go about it.
How you're structuring your journey through data
science and what [inaudible 00:15:50] you. So, to kick
us off could you, please, give us a quick overview? Who
is Sasha Prokhorova and how, what are you doing in
San Francisco?
Sasha Prokhorova: I'm currently a student at San Francisco State
University, pursuing my undergraduate degree in
electrical engineering. Originally I'm from Russia,
[inaudible 00:16:12]. That's where I obtained my first
degree in linguistics.
Kirill Eremenko: So why did you just jump from linguistics to electrical
engineering? That's a radical shift. That's like going
from South Pole to the North Pole.
Sasha Prokhorova: I do agree. I've always been interested in languages
while growing up, but also, when I grew older, I
haven't always been good in math. Not at least until
my early twenties. That's when I feel the gears really
shifted somehow because I noticed a lot of people say
that math is not really their thing because I think it
takes a certain age to be able to appreciate certain
mathematical concepts because of how abstract they
are and I'm inclined to believe that this is what
happened to me as well.
Kirill Eremenko: Okay, gotcha. That's very interesting and why data
science then? So electrical engineering, yeah, I
understand, but how is data science related to
electrical engineering and how are you leveraging it?
Sasha Prokhorova: First of all, we're living in a world that's drowning in
data, way more data than we can surmainly process.
I'm a firm believer that it's very important to have
certain data science and [inaudible 00:17:29] analytic
skills regardless of the industry you're in and in order
to maintain the edge in the competitive nature of
today's world. It's just impossible to, it's very
important to acquire those skills, at least, to any level.
Kirill Eremenko: Gotcha, gotcha. But is that just for technical
professions like electrical engineering or would you say
that's for management consultants and for, I don't
know, somebody running a bakery store or for
somebody who has a, who has a little tourism office?
Do you think it's important to have data skills for
anybody in this world?
Sasha Prokhorova: Well, yes, of course. We all produce data whether we
want it or not and our customers do produce data as
well, regardless of the industry we're in. If we are
bakers or management consultants, we all use and
produce data products to one extent or another.
Kirill Eremenko: Yeah, okay. I would totally agree with that. I think
some level data acumen or data knowledge is
necessary for anybody. But in your case, so electrical
engineering, data science, are you planning on moving
from electrical engineering completely to data science
or are you planning to integrate the two and have a
career that combines the two together?
Sasha Prokhorova: I don't believe I'm gonna move away from electrical
engineering. I just enjoy this industry way too much.
Currently, I'm working on a project of analog
integrated circuit designs and I'm having a great time.
But I do want to improve my data science skills and
knowledge and I'm currently trying to teach myself
some [inaudible 00:19:23] because it's just another
passion of mine. Something that I enjoy to the great
extent. I started going to some extracurricular classes
outside of school in San Francisco. Thankfully, the
data community is very strong in San Francisco and
they offer us has a lot of resources to improve our
skills and perfect ourselves. So, there's definitely a lot
of things that you can explore and try. Such things as
boot camps or evening workshops that you can just
explore before you commit to the full time course. It's a
great way to discover your passions and full-time
course. It's a great way to discover your passions and
interests and maybe even hidden skills and talents.
Who knows?
Kirill Eremenko: Mm-hmm (affirmative). Gotcha. Gotcha. And now let's
think about the other way around. So you already
mentioned how you're going to use data science. Why
are you going into data science now and like how that
can help augment your career and take it to the next
level. And in fact, how that could help anybody. But
tell us the other way around, like how does your
existing background help you be successful. As you
mentioned, you have quite a diverse background, with
linguistics and electrical engineering. How do you
leverage your background to be successful or be more
successful in data science?
Sasha Prokhorova: Well, actually I just started reading your book called
Confident Data Skills, which I find an incredibly
interesting read. And one of my favorite portions of it
would be quote that data science is one of those skills
that benefits from having experience in a different
field.
Kirill Eremenko: Mm-hmm (affirmative).
Sasha Prokhorova: Such as linguistics in my case. Or history or
management or consulting. I have very unusual
background for other young professionals who are
working in my industry. And I also have a very
unusual angle that I approach problems, which also
gives non-standard solutions.
Kirill Eremenko: Awesome, well tell us a bit about that angle. How
would you describe the angle at which you approach
data science problems? Very interesting.
Sasha Prokhorova: I would believe it's my ability to approach
unstructured data due to my data in linguistics. And
it's just my ability to read certain connotations that
maybe a non-linguist would not identify right away.
Kirill Eremenko: Mm-hmm (affirmative). Okay, that's very true. Very,
very interesting as well. So you're combining your
linguistics unstructured data skills with... And what'd
you get from electrical engineering? What kind of
mindset or thinking do you leverage from that field?
Sasha Prokhorova: Mathematical background. It definitely implies a lot of
structure, a lot of logic and a lot of discipline.
Kirill Eremenko: Okay. Gotcha. All right. So tell us then how do you go
about learning data science? Like are you taking
courses? Are you reading books? All right you
mentioned like you're reading my book, which thank
you very much. I'm very humbled to hear that you're
enjoying it. What are your main points of contact with
data science?
Sasha Prokhorova: I would definitely recommend couple of good books.
One of them would be Learning Python the Hard Way.
And there's also Statistics for Data Scientists. It's
really well written and not a difficult read at all. But
also use a lot of online resources, such as Code
Academy and DataCamp. There is a lot of very good
interactive exercises. And also I'm learning a lot of
MATLAB because my school requires it. It's part of the
curriculum for electrical engineers. And I recently
discovered that you can do data analytics and machine
learning in MATLAB, which made me even more
excited. I can use my engineering background and just
learn a couple new skills in MATLAB and I would be
able to use this incredible and powerful tool for data
analytics.
Kirill Eremenko: Mm-hmm (affirmative). Yeah, wow. That's a very good
recommendation. So you started learning data science
with Python. Is that correct?
Sasha Prokhorova: Yes. I was inspired by Craig Sakuma. He's one of the
instructors in General Assembly. It's a school in
downtown San Francisco. He taught me some Python
and some SQL. And he was actually one of those
mentors who made me believe that I can do it. I can
program. I can learn coding. The way he taught Python
and especially SQL, it totally made sense to me. He
basically did what Ben Taylor suggested to do all
along. Find the project that's exciting and important to
you. He did it based on the music. We were analyzing
his iTunes playlist in SQL. Not necessarily just for the
genre or for the length of the songs, but for instance,
how many songs does Craig have in his playlist that
are love songs? And also what signifies a love song? Is
it the word love, hug, kiss, or could they be used in
any sarcastic contexts? That could be, that's one of the
tougher projects for machine learning, too.
Kirill Eremenko: Okay. That's a very interesting project. When you were
at the conference at [inaudible 00:25:03], did you
attend Sinan Ozdemir's talk?
Sasha Prokhorova: Yes. Yes, I did.
Kirill Eremenko: Because Sinan is also an instructor at General
Assembly. Or maybe he was, but he definitely spent
quite a lot of time at General Assembly. And I just... In
San Francisco as far as I remember. Did you know
that about him?
Sasha Prokhorova: No, actually I did not. I cannot believe I missed out on
that.
Kirill Eremenko: Yeah. Yeah, well, there you go. Yeah, I heard they have
some very nice courses there. Okay. So do you attend
like the General Assembly events in San Francisco?
Sasha Prokhorova: I do frequently. That's something that I enjoy doing
after my regular classes at school. I would say spend
the whole day at campus at San Francisco State,
attending lectures and labs, I would spend some time
in the library. But then in the evening, I would find
something that's interesting and appealing to me that
sounds like I might enjoy and I just go check it out.
And General Assembly... And I just have fun meeting
different people and learning new things.
Kirill Eremenko: That's very cool. And are those... How are you... In
terms of technical complexity, how would you describe
the General Assembly classes? Just for like listeners
out there. Because General Assembly's not just San
Francisco, it's all... I think it's nationwide for the US,
maybe somebody else might want to attend one of
these. Like would you recommend it for beginners or
advanced data scientists? What kind of level do they
have?
Sasha Prokhorova: You know, that's the beauty of this place. It's tailored
for very diverse crowd. It works for very complete
beginner. Even for someone who is just very curious
about data science or machine learning. They can just
attend an evening workshop and just get the gist of it
and decide if it's right for them or not. And they have
more advanced programs as well. Such as bootcamps
and more full-time courses. So yeah, it's an amazing,
amazing resource.
Kirill Eremenko: Mm-hmm (affirmative). Okay. Awesome. All right. Well
tell us a bit more about your... You know, you
mentioned you learned Python already. And how did
you find learning Python? Like obviously everybody's
background is different, and you had some experience,
I'm assuming you had experience in MATLAB before
Python? How did you find Python after MATLAB?
Sasha Prokhorova: I enjoy [inaudible 00:27:23] a lot because it truly made
sense to me. It was very similar to MATLAB and the
search and Python syntax structures. They echoed
MATLAB in my brain.
Kirill Eremenko: Okay. Gotcha. And is there any other tools that you're
looking forward to learning sometime soon?
Sasha Prokhorova: Tableau. That's actually one of my good friends, and
one of my mentors who I met at the Open Data Science
conference last year. His name is Pratyush [inaudible
00:27:52]. He suggested that I should learn Tableau as
a first step in my data analytics journey, and just
create a project in Tableau and showcase it.
Kirill Eremenko: Yup.
Sasha Prokhorova: Because Tableau is known to be a very flexible and
eloquent, and yet very powerful tool. And I think it
could be a good starting point for any aspiring data
scientist or analyst.
Kirill Eremenko: Wow. Definitely. I really like Tableau, that's kind of like
where... I think I started my data science journey from
SQL, then I moved to Tableau, then came R and
Python. Everybody has their own way. But, yeah. It's
good to always kind of be looking forward to the next
step, the next thing that you're gonna be learning. So
tell me this Sasha, do you take courses on Udemy?
Just out of curiosity?
Sasha Prokhorova: Yes, I do. I actually took your and Hadelin's course
about data science careers. I downloaded a couple of
courses about Python and I'm actually very excited to
embark on that journey. Yeah-
Kirill Eremenko: Awesome. And I'm assuming, well from what you told
me, that you listen to the Super Data Science podcast
as well?
Sasha Prokhorova: Yes. It's actually one of my favorite podcasts. I
discovered it when I was commuting to my industrial
engineering internship in [inaudible 00:29:14] and
yeah, I just came across it. And I was so grateful and
lucky that this resource fell on my lap. Because I was
actually looking forward to my commute to work so
that I could listen to the podcast.
Kirill Eremenko: That's awesome. Thank you. Thank you for the
comment. And tell me, how long have you been
exploring data science for so far?
Sasha Prokhorova: I wanna say for about a year and a half.
Kirill Eremenko: Yeah and a half? Okay, so the reason why I'm asking
all these questions is because I'm trying to
understand... Or actually I just want to show to our
listeners what passion means. What passion looks
like. So as you can see, Sasha is reading books on
data science, listening to podcasts, taking courses on
Udemy and Code Academy and DataCamp. She's
attending conferences, not just DataScienceGO, but
she's also been to ODSC. She's attending the General
Assembly occasionally when she feels like doing
something fun after a hard day at University. Winding
down with some data science at General Assembly.
And I'm sure there's lots more other things that you do
in this space. You follow people like Randy Lau on
Linkedin and you find ways to get in touch with Ben
Taylor, or maybe meeting him at a conference and
asking him about some advice. So you're getting
mentors directly or indirectly.
Kirill Eremenko: So as we can all see, like you, this, I wanna just show
to our listeners, especially those who are starting out
or those who want to like propel their career and you
might be finding that your career's not really going
where you want it to. Well, as Ben Taylor described in
his talk, you've gotta be passionate about something.
And this is what passion looks like. To me this is what
passion looks like, these are the indicators of passion.
Sasha is definitely a person who is passionate about
the field of data science. Because otherwise she
wouldn't be doing all this. Sasha would you agree that
you're passionate about data science?
Sasha Prokhorova: Absolutely. I would say passion has a power to move
mountains if you are determined enough.
Kirill Eremenko: Mm-hmm (affirmative). Definitely. Definitely. And now
you're on this podcast. And I don't think that's a
coincidence. Like I... Probably when we were there we
didn't talk for long, but already just by your
excitement and energy that you came into that
conversation with I could feel your, you know passion
it kind of like translates itself. And so therefore when
somebody who's passionate, like maybe Sasha in your
case, when you go for an interview in data science,
you're gonna like in a 30 minute interview, the
recruiters or data science manager, they will feel that
from you as well. Just like how I felt it. And hence, it
will be so easy for you to get any kind of career that
you want. And people who don't feel it, that is kind of
like they're going to be missing out.
Kirill Eremenko: And that's for our listeners out there, once again it
doesn't matter if you're just starting out into data
science or you're already an expert in data science,
you wanna position yourself like that. You wanna be
the person that's emanates this energy, this passion or
bordering on the level of obsession, that people will
come to you with job offers. So Ben Taylor had this
example in his talk that there was a group of students
that he was talking to and all of them were like you
know I would love a job in data science, but it's so
hard to find one. And among them there's this one
student out of like maybe seven students. Among them
there's this one student who had all the job offers
because all the employers in the area or all the
companies that knew about this group, they knew that
this guy was super passionate and they could feel it
from when he was sharing online, how he was talking,
what he was doing. And you want to be in that
position. You want to be getting all the job offers.
Would you agree with that, Sasha?
Sasha Prokhorova: Absolutely.
Kirill Eremenko: Yeah.
Sasha Prokhorova: It's important to love what you do and have good work
ethic. And just keep trying and trying and trying
without being afraid of failure. Because failure's just a
natural part of the learning process and it's inevitable.
And I think, actually as you said during the
conference, that we learn a lot about success, but we
also learn ten times more from failure. Because as long
as it's important to know what to do, from failure you
actually learn exactly what not to do.
Kirill Eremenko: That's a great way of putting it. Okay, speaking of
failures, tell us a bit about, what is... Or let's talk
about your failures. What would you say has been like
your biggest failure, that you've learned from the most,
in this pursuit of data science and technology and
data and career some.... Attached to data.
Sasha Prokhorova: Well I wouldn't necessarily call it a failure yet, because
I'm just so new to this industry. I haven't even entered
yet. I would call it a temporary lack of result.
Kirill Eremenko: Mm-hmm (affirmative).
Sasha Prokhorova: Because it's also important to know how to approach
recruiters correctly because this field is so competitive
and it's so cut throat. And recruiters, both in Linkedin
and in real life, they're so overwhelmed by the volume
of applications they received. So I've applied to
probably hundreds of positions that are relevant to my
field and I either received either thank you but no
thank you or no response whatsoever. But I don't let it
discourage me, I just keep trying. So basically, short
answer to your question, my main failure would be not
getting an entry level position yet, since I'm still at
school.
Kirill Eremenko: Yup. [crosstalk 00:35:10]
Sasha Prokhorova: But my main takeaway from all this job hunt and the
conference would be for recruiting managers correctly.
I have the theory that I call what keeps you up at
night. You would ask the manager what are the main
challenges that your company faces nowadays and
what can I do to help you to solve those problems? To
improve your company and to achieve the goal by the
end of the year that you want to. What can I help you
with to help us both succeed?
Kirill Eremenko: Mm-hmm (affirmative). Okay. All right. So, I've got a
few comments here. So first one, I would like to
comment that I wouldn't agree that it is a cut throat
field, and I'll explain why. Because when I was a
consultant in Deloitte, right? And I know what cut
throat means and what cut throat looks like. And that
is like a completely different story when you are, when
people who you're working side by side with... I'm not
talking about this about specifically at Deloitte, so
don't wanna get anybody in trouble or anything like
that. But just I've seen the world of consulting, and
that is cut throat, right? Like when people are, like you
kind of like think they're friends, and then there's
promotions in question and you have this two year
policy to... Like you're either up or out within two
years. You either get a promotion or it's implied that
you leave the company because you're not good
enough. And you know, in that kind of environment,
where everybody's competing with each other, that's
what I define as cut throat.
Kirill Eremenko: And data science, I think data scientists as a
community are much stronger. Like I wouldn't call
consultants as a community, like I'm sure they are
communities in consulting that are fantastic, but
overall in general it is more cut throat. Whereas in
data science, everybody wants to help everybody.
Everybody's sharing their code, everybody's
commenting on each other's mistakes. There's plenty
of resources like Quora and Stack Overflow and Kaggle
and wherever you ask your question, you get answers
very quickly. I would say it's a more communal effort.
But I do agree with you in the sense that, the fact that
there is so much, like there's a massive demand for
data sciences, but there's an overwhelming supply.
There's so many people that have gotten into data
science just for the sexiest job of the twenty-first
century or the massive salaries and so on, that are
there maybe for the wrong reasons. Or that are... You
know recruiters have so much to choose from, and in
that sense yes, it can be very difficult to get those
applications and job positions. So, in that sense,
disagree that it's cut throat. I would say that
terminology is different, it's just that it's overwhelming
supply at the moment.
Kirill Eremenko: On the other hand, what I wanted to say is, do you
mind if I give you a bit of advice in terms of how you
approach your career? And I think that it would be
helpful for our listeners as well.
Sasha Prokhorova: Please do, I would love that.
Kirill Eremenko: All right. So what I would say in this case is what
you're doing, I would say what you're doing wrong and
what a lot of people are doing wrong, is they're going
for the recruiters. Yes, inevitably you're going to send
hundreds of job applications and you're going to get
refusals, you're gonna get people turning you down.
And it is not a reflection of your skills or passion. Like
we already established on this podcast already that
you're definitely passionate about data science, you're
doing so many things, you're learning. You're gonna go
a very long way in this field. Like I can already tell that
you have a very bright future.
Kirill Eremenko: The question is, how do you people, as you said,
there's so many job offers or job applications that
recruiters get that they get like for every offer, for every
job posting they get maybe a thousand, I don't know a
couple hundred job applications. And it is physically
impossible to go through them. So no matter how great
you are, if you're going through the standard pipeline,
standard process, you will find that you are, they
might just not see your application in the first place.
Like if they were to see it, then you'll stand out to
them. But if they don't see it, it's never gonna stand
out. And moreover, as they say about 70 to 80 percent,
not just as they say, studies have shown that 70 to 80
percent of jobs are filled or job postings are filled
behind the scenes. They're never actually posted
online for everybody to see. What we see online is all
these jobs offers or job positions that recruiters and
managers need to fill. That's just 30 percent of the
whole job market. Most of the jobs get filled through
referrals, through managers going out there before
posting a job and just like looking for somebody
through friends of friends, through people in your
network on Linkedin and stuff like that. So, first step
is, we only see 30 percent of the demand for data
science. We only see 30% of the demand for data
scientists and moreover, for every job, there's
hundreds of applications and therefore nobody sees
your application. So, it's a losing game. You're playing
a losing game and some people turn to get up numbers
and they send a thousand applications and maybe one
or two succeeds.
Kirill Eremenko: That's not the opposition that you want yourself to be
in. Right? You don't wanna be scavenging for jobs and
only getting the one where the manager did notice your
application and therefore, you're just picking out of
one or two jobs that might not ultimately be the best
job for you, but that's all you have to choose from. You
wanna flip the table. You want to be in the ocean of
people applying for jobs, an ocean of applicants or
data science professionals. You want to be like a
shining star.
Kirill Eremenko: You want to be something that stands out, like if you
look at an ocean in the darkness of the night and
there's nothing there, it just looks black But if there's
a ship sailing from left to right, you will see the ship
right away, right? It stands out. So you wanna be that
shining star. And how do you get to that level? How do
you become the shining star?
Kirill Eremenko: Well, it's actually ... there's nothing difficult about
that. You just have to start building your brand online.
You have to start making some noise. You have to
start making some ripples in the water so that you do
attract attention, because if you're doing same thing as
everyone else is doing, there's no way you're gonna
stand out.
Kirill Eremenko: For instance, that's what I did, and I did this a long
time ago when I was, you know, when was this? 2014,
so four years ago when I was leaving Deloitte and I
decided, I want a job. I don't wanna be in consulting
anymore. I want a job. And one way I could've gone
about it, and I tried to do it, but then I didn't have
time, because I was still working at Deloitte. And I
looked around, and one way I could do it is apply for
jobs in data science, but then I realized that it's taking
too much time. I'm way to perfectionist to just send
out a standard template resume to all these jobs. I
wanted to tailor my resume to every single position,
write a cover letter. That was taking me hours for every
application, and that was not sustainable, and not
scalable. So I couldn't-
Sasha Prokhorova: Yeah. I'm guilty of that too.
Kirill Eremenko: Yeah, and so I couldn't turn that into a numbers
game. I couldn't send thousand applications out at
once, because I knew that I'm too perfectionist for
that. So instead what I did was, alright, let's flip the
game. Let's flip the table, and instead I'm just gonna
start posting on LinkedIn. Not even huge stuff. Not
even [inaudible 00:42:33] writes an article, which
takes a few weeks to write. I just started reading,
finding stuff online that is relevant to data science and
technology, reading it, commenting on it in like one or
two lines of comments, and then re-posting it on
LinkedIn, saying, hey guys, I found this article. I found
it interesting. This is what I thought about it. And I
think it's controversial. I think, I agree, or I disagree,
it's my opinion. And I would post that and I actually
automated the posts. I would read all those in the
weekend and then I would post them, get a tool like
Hootsuite and post them throughout the week. You
know you gotta post it like three times on Monday, or
Tuesday, Wednesday, and Friday. Or Thursday when
people actually read that stuff.
Kirill Eremenko: And within six weeks, magic happened. I started
getting recruiters checking my profile, I started getting
managers, and within six weeks I got three job offers.
I'm not making this up, I had three jobs offers within
six weeks for a very basic LinkedIn profile with only a
couple of years experience in the field. All I did was
just start making some noise and I didn't write my
own class of articles, I just commented on stuff and I
got a job offer from a pension fund which is in
Australia called [inaudible 00:43:43] fund, in the city
that I lived in and I got two job offers from banks in
Sydney. From very large banks, I think both of them,
or one of them was one of the big four banks in
Sydney. I actually went through the interview process
with I believe all three companies and then two of
them, the third one, I just didn't go to the final
interview stages. Two of them gave me job offers which
one of them I picked and I went to it and I worked
there.
Kirill Eremenko: In essence, all of them were almost double the salary I
was making at Deloitte. So, not only I got job offers,
not only I got double my salary, but I actually didn't
have to do much. I didn't have to apply for any jobs
myself. They just came to me. Right? So and now it's
been four years later, the demand for data scientists
has skyrocketed, the applicants - there's still an ocean.
It's a bigger ocean, but it's not an ocean - not many
people doing much about standing out. Still an ocean
of applicants. But the demand has skyrocketed, so it's
so much easier to stand out now. All you have to do is
make some noise, post some articles, plus you could
write about what you learned in general assembly,
write a little article about how you went into data
science, what you learned there. You don't even need a
blog, you just write those in LinkedIn [inaudible
00:45:00] share them there.
Kirill Eremenko: You could write about what you're reading in a book,
what you're taking in a course. You know, write up a
couple of those things, share this podcast episode that
you've been on. I have no doubt that within, by the
start of next year, by the start of 2019, you will have
so much attention. If on LinkedIn, if you get premium
you can see who [inaudible 00:45:21] who is visiting
your profile, who is seeing what you're doing, what the
company they're from, what positions they are. You
will see slowly managers will start popping up,
recruiters will start popping up, and then the job offers
will start coming. And that's all it takes. That's my
thoughts on this.
Kirill Eremenko: What do you think?
Sasha Prokhorova: I completely agree. I believe blogging is the new CV as
mentioned by Andy Parker, a medium whom I follow.
And it just important to generate quality contents, and
just put yourself out there.
Kirill Eremenko: Okay, my question is why, if you believe that, why are
you not doing that? Was something preventing you
from doing that?
Sasha Prokhorova: I'm trying to accumulate more skills and more
knowledge that I can share with people.
Kirill Eremenko: Oh my, this is the typical issue. Why! You hear this all
the time. This is fear that I am not enough. This is fear
that I'm not good enough. You've been in this one and
a half years. You can have so many people. There is
literally 100s of thousands, as we've discussed at the
conference, there's a shortage of 173,000 data
scientists nationwide in the US right now. There's so
many people. There's 100s of thousands of people
going into this field. You're one and a half years of
experience of learning data science is golden to tens, if
not 100s of thousands of people.
Kirill Eremenko: You can start now. Just take the first step, write the
first article. Make it, or just share something. You will
see how many you have helped. And even if you help
one person, that's already a massive step and trust
me, you don't need to be an expert in this field to be
able to share your experience and help people.
Sasha Prokhorova: Maybe all I need is just to begin. I just need to find a
project that I am passionate about, and just start
writing, start exploring, and just trying to find out
what I'm capable of.
Kirill Eremenko: Okay. How about, do you want to make a commitment
on this podcast publicly? Like Rico says -
Kirill Eremenko: What's it called? What is the term, a radical
commitment? Or something like that?
Sasha Prokhorova: A reckless commitment.
Kirill Eremenko: Reckless commitment. How about we do one of those
right now.
Sasha Prokhorova: Absolutely. I was actually quite enamored by his words
that you can be an expert in something in three
months if you commit to it. Let's say my commitment
for the next three months would be finding a data set
on let's say, [inaudible 00:47:53] or some resource like
that, and just start working with it, starting to look for
patterns and see what I can make of it.
Kirill Eremenko: Love it. I totally love that. I think we will lock that
commitment in, but I think you can do better. So, do
you have any exams in the next week?
Sasha Prokhorova: Yes, actually.
Kirill Eremenko: How many exams do you have?
Sasha Prokhorova: I'm kinda half-way through my midterms, so on
Monday I have my integrated circuit design class,
where I have to analyze the performance of certain
MOSFET transistors.
Kirill Eremenko: And then after Monday?
Sasha Prokhorova: I have a communications systems and I also have a
power systems.
Kirill Eremenko: Okay. Cool. So, do you think you'll be able to find
three hours of free time until Thursday next week?
Sasha Prokhorova: Absolutely! I think I'm gonna break it into the
increments of 30 minutes each day, which would bring
me to six days of week on working on the project,
without it taking away too much from my course work.
I'm convinced I can do that.
Kirill Eremenko: Awesome. Okay, so, the new commitment that I'm
offering to you right now on the podcast is that - so
this session, we're recording this today on Friday the
19th of October, and this session is going to go live on
evening of Wednesday the 24th of October. So that is
five days away. My challenge to you is can you write a
500 words article that you're gonna share on LinkedIn
by Wednesday, and then our listeners will be able to
check, because once you write it, you will send us the
link, or you send me the link, and I will include it in
the show notes. So as soon as this session is live, so
for our listeners, when you're hearing this, this session
is live, on iTunes or SoundCloud, wherever you're
listening to it, and that means that by now Sasha has
finished writing her first 500 word article and it's live
on LinkedIn and you can go to the show notes and
check it out there. So the show notes, I'll announce
where they are, and [inaudible 00:50:05]. You can go
to show notes, click on it and read it.
Kirill Eremenko: How does that sound to you Sasha?
Sasha Prokhorova: It sounds really good. I'm very excited, yes, let's make
complex, simple.
Kirill Eremenko: Nice, nice. Very good. So, we're going to do that and
that's a good way - and then you can then e-mail Rico,
and say, hey Rico I did your reckless commitment
thing and this is what I came up with. And that's a
good way to get - sometimes we need to kickstart,
right? So, we need somebody or something external to
force us to actually do something about our careers
and lives, and this is going to help you kickstart into
the process and then hopefully, once you've written
the first article, and you've seen how many people
you've helped and how easy it is - then you will get
into the mood for it and maybe you'll start writing one
per month or two per month, and that is the way, I
think for you and many people in this field, that is the
way to cause those ripples on the water so that you
will be seen by recruiters and managers.
Kirill Eremenko: Sounds good?
Sasha Prokhorova: Yeah! I think, it sounds great. I think we all need to be
pushed out of our comfort zone every now and then,
and this is kind of what happened to me today, and
especially seen Rico's and yours, and using it
[inaudible 00:51:22] as knowledge and dedication.
Those [inaudible 00:51:24] are really and truly
contagious, and inspiring to me.
Kirill Eremenko: Fantastic. And I want to actually extend this invitation
further to our listeners. If you are stuck in the same
boat as Sasha and you know you've been applying for
jobs, especially if you've been applying for jobs and
with no success, then I challenge you to take one week
- so this podcast is going live on the 24th, evening,
24th of October, so one week until the first of
November, to write something. Doesn't matter how
long it is, aim for 500 words, but even if you do 200,
that's enough. And, share it on LinkedIn. Write, and
then see what happens. See how you feel, how long it
takes you, and in fact in order to make this even more
fun, send your link to your article to podcast at
SuperDataScience.com, and we will pick the best one
and we will reshare it on our LinkedIn. So we will pick
the best one that you guys write up, and we will
reshare on our LinkedIn with 25,000 plus followers. So
you can actually impact a lot of people.
Kirill Eremenko: But it has to be done by the 31st of October, so all
submissions need to be in first of November. So,
there's my challenge out to you guys, and that will
help you kickstart your work career and if you already
have a career as well, if you are already a successful
data scientist doing plenty of work and your happy
with everything, it's a great way to give back to the
community and I also encourage you to put this bid
into this challenge.
Kirill Eremenko: See Sasha what you did! You started a whole thing.
Sasha Prokhorova: Hashtag homework challenge!
Kirill Eremenko: Hashtag - yeah, let's call it hashtag SDS homework
challenge. One word. Okay, awesome. That is really
fun so I'm going to actually put a reminder for myself
to check the submissions. What else? Tell us what
else. We're slowly coming to a wrap up of this podcast,
what are some thoughts that you would like to share
with your fellow data scientists, maybe people you met
at DataScienceGO, maybe - just people who are
listening to this podcast, or getting into the field of
data science or moving from another field into data
science. What is something that you would like to
share [inaudible 00:53:39]?
Sasha Prokhorova: I would say don't stop exploring and don't be afraid to
discover new interests.
Kirill Eremenko: Mm-hmm (affirmative). Curiosity, right? Is the
[crosstalk 00:53:50]
Sasha Prokhorova: Curiosity - absolutely.
Kirill Eremenko: Yeah.
Sasha Prokhorova: Curiosity and passion and thirst for knowledge, for
constant learning. Next stop learning.
Kirill Eremenko: Yeah that's very true. Very true. I find curiosity - you
know [inaudible 00:54:04] obsessions? I find curiosity
my obsession. I find that sometimes I - for instance,
let's say I'm cooking something, I don't know, a pasta
or some beans or something like that, and then I know
that the recipe says do this, but then sometimes I just
have this idea of what happens if I do this? What
happens if I put the ingredients in the wrong order?
What happens if I add this ingredient that's not in the
recipe, or what happens if I replace this with that. And
sometimes, it's just such a burning desire to explore
what will happen that I just like, what if? And I do it.
And then it's either, most of the time it's a failed
result, to be fair, but sometimes something epic comes
out of it. It's not just in cooking, it's pretty much in
anything that I do. I always - as soon as I have this
question in my head, what if? I don't let it slip away.
Very rarely have I let it slip away. I'm always, okay,
let's do it.
Kirill Eremenko: Screw it, like Richard Branson says, screw it, let's do
it, let's see what happens. And I think that's curiosity,
right, in data science you gotta be the same. You gotta
be like, what if I write logistic regression in this way,
what if I apply this data set, what if I worked on this
project, what if I read this book and so on. What will
happen? And don't let that slip away. As soon as you
have the what if, there's always gonna be another voice
saying, I'm too tired, I'm too lazy, I really know that
this other method is gonna work. Well nothing new
comes out of doing the same things the same old way.
You gotta try new things, and that's when you break
boundaries, whether it's in science, in exploring new
fields, or whether it's in your career and your personal
life, in general and things that you are capable of
doing.
Sasha Prokhorova: Yeah absolutely.
Kirill Eremenko: Awesome. Okay well, Sasha, thank you so much for
such an inspiring session. I had a massive pleasure
talking to you, I'm sure lots of people learned from
this. Before I let you go, could you let us know where
our listeners can get in touch, contact you, follow you,
learn more about your career and see where it takes
you?
Sasha Prokhorova: Feel free to follow me on LinkedIn, it's Aleksandra
Sasha Prokhorovaa. Sasha in the parenthesis, because
Sasha is short for Aleksandra in Russian. Yeah the
last name is Prokhorovaa.
Kirill Eremenko: Wonderful, okay. So, well also include the URL to your
LinkedIn in the show notes for all listeners to catch up
on there. On that notes, once again, thanks so much
for coming on this show, best of luck with your career.
Let's stay in touch and I am looking forward to reading
your article for Wednesday next week.
Sasha Prokhorova: Thank you very much Kirill.
Kirill Eremenko: Alright, take care.
Kirill Eremenko: So there we go, that was our chat with Sasha. I hope
you enjoyed it. You can get the show notes and check
if Sasha has completed her challenge successfully at
www.superdatascience.com/203. There you'll get all
the show notes and everything from this podcast, all of
the things that we've mentioned.
Kirill Eremenko: Another thing I wanted to outline today, before we
finish off, is that during this session we talked about
DataScienceGO 2018 quite a lot. As you could see, and
feel, and hear, Sasha had an amazing time. There was
plenty of speakers there, and lots of things to learn. So
if you missed out on DataScienceGO 2018, or if you
attended and missed out on certain sessions, because
we did have two rooms in parallel, so if you missed out
on certain sessions, then I have some great news for
you. You can get the recordings from DataScienceGO
2018 and keep them for life, today. You can go to
DataScienceGO, www.datasciencego.com/recordings,
and you will find all of the sessions there. You'll be
able to purchase the whole package and keep it for life,
and revisit any talks that you loved if you were there,
any talks that you missed, if you weren't there, and get
all the value of it. We recorded every session with
professional camera crew, so the quality is outstanding
and you get the full package both Saturday and
Sunday included.
Kirill Eremenko: So make sure to check out
datasciencego.com/recordings, if you want to relive
this experience or get all of the value from our
speakers that our attendees got. Of course you won't
be able to get the networking that is something you get
only by being there, but at least you can get all of the
value that our speakers were there to share in terms of
their talk, in terms of the things that they prepared for
this conference.
Kirill Eremenko: So, highly recommend checking it out, it's
datasciencego.com/recordings, you can find it there,
and on that note, thank you so much for being here,
for being part of the SuperDataScience podcast today,
spending this hour with us, with me and Sasha, and I
look forward to seeing you back here next time. And
until then, happy analyzing!