A2 初級 美國腔 276 分類 收藏
>> Announcer: Live from
Stanford University

in Palo Alto, California, it's theCUBE.
Covering Women in Data
Science Conference 2018.

Brought to you by Stanford.
>> Welcome back to theCUBE, we are live at
Stanford University for the third annual
Women in Data Science
Conference, hashtag WiDS2018.

Participate in the conversation
and you're going to see people
at WiDS events in over 177

regions in over 53 countries.
This even is aiming to reach about
100,000 people in the next couple of days,
which in its third year is remarkable.
It's aimed at inspiring and
educating data scientists

worldwide and of course
supporting females in the field.

It's also got keynotes, technical
vision tracks, and a career panel.
And we're excited to
welcome back to theCUBE,

a cube alumni, Ziya Ma,
the Vice President of

Software and Services Group and the
Director of Big Data
Technologies at Intel.

Ziya, welcome back to theCube.
>> Thanks for having me, Lisa.
>> You have been, this is your
first time coming to a WiDS

event in person and your first year here.
You are on the career panel.
>> Yes.
>> That's pretty cool.

Tell us about, you just came from
that career panel, tell us about that.
What were some of the
things that excited you?

What are some of the things that
surprised you in what
you heard at that panel?

>> So I think one thing
that was really exciting

is to see the passion from the audience,
so many women
excited with data science.
And it was the future of
what data science can bring.

That's the most exciting part.
And also, it's very
exciting to get connected

with so many women professionals.
And in terms of,
you know, surprise?
I think it's a good surprise to see
so much advancement in women
development in data science.

Comparing where we are and
where we were two years ago,

it's great to see so many
woman speakers and leaders

talking about their work
in the data science space,

applying data science to
solve real business problems,

to solve transportation problems,
to solve education, healthcare problems.
I think that's the happy
surprise, you know,

the fast advancement with woman
development in this field.

>> What were some of the
things that you shared,

maybe recommendations or advice.
You've been in industry for a long time.
You've been at Intel
for quite a long time.

What were some of the things that you felt
important to share with the audience,
those in-person here at Stanford
which is about 400 plus,

and those watching the live stream?
>> Yeah, you know, Lisa,
I provide career coaching

actually for many women professionals
at Intel and also from the industry.
And a lot of them expressed an interest
of getting into a data science field.
And they ask me, what is the skillset
that I need to develop in
order to get into this field?

I think first, you need to ask yourself,
what kind of job you want
to get into in this field.

You know, there are marketing
jobs, there are sales jobs.

And even for technical jobs,
there are data engineering

type of jobs, data visualization,
statistician, data
science, or AI engineer,

machine learning, deep learning engineer.
So you have to ask
yourself, what kind of job

you want to move to and then
assess your skillset gap.

And work to close that gap.
Another advice I give to
many woman professionals

is that data science appears
to have a high bar today.

And it may be too significant a jump
to move from where you are
to a data science field.

You may want to move to
adjacent field first.

And to have a sense of what is it
like to work in the data science field
and also have more insights
with what's going on.

And then, to better prepare you
for eventually moving into this field.
>> Great advice and I think
one of the things that jumped

out at me was you talked about skillsets.
And we often hear a lot of
the technical skills, right,

that are essential for a data scientist.
But there's also softer
skills, maybe it's more

left brain, right brain,
creativity, empathy, communication.

Tell me, in your ascension
to now the VP level at Intel,

what are some of the other skills besides
the technical skills that you find
as data science as a field grows
and infiltrates everything,

what are some of those softer skills
that you think are really advantageous?
>> Great question.
I think openness and collaboration
are very important soft skills.
Because as a data scientist, you need to
work with data engineering teams.
Because as a data scientist, you extract
business insights from the data.
But then you cannot work alone.
You have to work with
the data engineering team

who prepares the data infrastructure,
stores, and manages the data very
efficiently for you to consume.
You also have to work with domain experts.
Let's say if you are applying
data science solutions to

solve a real business problem,
let's say in a medical field.

You need to work with a domain
expert from the medical field

so that you can tailor your
solution towards, you know,

addressing some medical problems.
So you need to work
with that domain expert

who knows the business
operations and processes

in medical field really, really well.
So I think that's, you
know, collaboration is key.

And of course you also want to collaborate
maybe with academia and
open source community

where a lot of real
innovations are happening.

And you want to leverage
the latest technology

building blocks so that
you can accelerate your

data science application
or solution advancement.

So collaboration and openness are the key.
>> Openness is a great one.
I'm glad that you brought that up.
We had another guest on
talking about that earlier.

In terms of being open, one,
to not expecting, you know,
in the scientific method,

you go into it with a hypothesis
and you think you know
what you're going to find

or you want to know, I want to find this.
And you might not, and
being open to going,

okay, that's okay, I'm
going to course correct.

'Cause failure in this
sense is not a bad F word.

But also being open to other
opinions, other perspectives.

That seems to be kind of
a theme that we're hearing

more about today, it's be
willing to be open-minded.

>> You know, that's an
excellent point, Lisa.

You know, I can share one example.
When coming from an
engineering background,

when I first moved into this field,
we always had the assumption that
when we talk with your customers,
they must be looking for
something that's high performance.

So our initial discussion
with our customers

centered around Intel product lineup
that will give you the
highest of performance

for deep learning training
or for analytics solution.

But as we went deeper with the discussion,
we realized that's not what customers
are looking for in many cases.
The fact is that many
of them have collected

a massive amount of data over the years.
They have built analytics applications
and you add on top of that.
And so as the data
representations get more complex,

we want to extract more complex insights.
That's the time they want
to apply deep learning

but to the existing
application infrastructure.

So they're looking for something,
let's say deep learning
capability, that can be easily

integrated into the existing
analytics solutions stack,

into its existing infrastructure
and reuse its existing

infrastructure for
lower cost of ownership.

That's what they are looking for.
And high performance is just nice to have.
So once we are open-minded
to that learning,

that totally changed the conversation.
Actually, in the last couple of years,
we applied that learning and
we have collaborated with

top cloud service providers
like Amazon, Microsoft,

Google, and you know, Alibaba and Baidu
and a few others to deploy
Intel-based deep learning capabilities.
Libraries, frameworks, into
cloud so that, you know,

more businesses and
individuals can have access.

But again, it's that openness.
You truly need to understand
what is the problem

you are solving before simply
just selling a technology.

>> Absolutely, and that's one of the
best examples of openness that's obviously
in this case listening to customers.
We think we know the problem
that we need to solve

and they're telling you,
actually, it's not that.

It's a nice to have, and you go,
whoa, that changes everything!
And it also changes, sounds like,
the downstream collaboration
that Intel knew we need to have

in order to drive our business forward
and help our customers in every
industry do the same thing.

>> Exactly, exactly.
>> So a couple of things that I'd love
to get your perspective on
is the culture at Intel.

You've been there a long time.
What is that culture like in terms of
maybe fueling or being a
nice opportunity for bringing

in this diversity that we
so need in every industry?

>> Yeah, you know, one thing
I want to share, actually,

just now during the panel
discussion I shared this.

I said Intel will be the first
high tech company achieving

full representation of women
and under-represented minorities
by the end of this year.

>> Wow, by the end of 2018?
>> Yes, we pulled in our
timeline by two years.

Yes, we're well on track for this year.
>> Wow.
>> To achieve that.

And I personally, I like this quote from
Brian Krzanich, our CEO, that
if we want tech to define

the future, we must be
representative of that future.

So in the last few years now,
Intel has put great effort

into hiring and retention for diversity.
We also have put great
effort for inclusion.

We want to make sure our
employees, every one of them,

come to work, bring their
full selves for the value add.

We also invest in diverse entrepreneurs
through Intel capital initiatives.
And most importantly, we
also partner with academia,

universities, to build the
pipeline for tech sectors.

So we put a lot of effort
and we committed about $300 million
for closing the gap at the company
but also for the high tech sector.
So definitely we are very committed
to the diversity and inclusion.
But that doesn't mean that
we only focus on this.

And of course, we make sure
that our people are bringing

the right skillsets and we
bring the most qualified people,

you know, to do the job.
>> On the pipeline front, one
of the things I was reading

recently is some of the
challenges that organizations

that are going to, say,
college campuses to recruit,

some of the missteps they might be taking
in terms of if they're
trying to bring more females

info their organization in STEM roles,
don't staff a booth with men, right?
Or have the only females that
are at a recruitment event

be doing, handing out
swag, or taking names.

Obviously there's important
roles to be had everywhere.

But that was one of the
things that seems to be,

well what a simple thing to change.
Just flip the model so that
the pipeline, to your point,

is fueling really what
corporations like Intel want

to achieve so that that future is really
as inclusive and diverse as it should be.
The second thing that you
mentioned before we went live,

from an Intel perspective,
is you guys were challenged

on the talent acquisition front.
And so a few years ago, you started the
Women in Big Data Forum
to solve that problem.

Tell us about that and what
have you achieved so far?

>> Great question.
So you know, this is
three or four years ago.

And Intel, you know, because
I manage the big data

engineering organization within Intel,
and we are working to hire
some diversity talents.

So we opened some racks and
we look at our candidate pool.

There were very few
women, actually barely any

women in the candidate pool.
Again, yes, we always want to hire
the most qualified people, but
it also does not feel right

that when you don't even have any
diversity candidates in that pool.
Even though we exhausted
all possible options,

even tried to bring
the relevant diversity
candidates into the pool.

But it's very challenging.
So then we reached out to a
few industrial partners to see,

is Intel the only company
that had this problem

or you have the same problem?
It turned out everyone
had the same problem.

So yes, people value diversity,
they all see the value.

But it's very challenging
to have a successful

recruiting process for diversity.
That's the time the few
of us gathered together,

we said, maybe there is
something that we can do

to support a stronger woman
pipeline for future hiring.

And it may take a couple of
years, and it may take one year,

but unless we start doing something today,
we're going to talk about the
same problem two years from now.

>> Exactly.
>> So then with sponsorship
from our executive team,

Doug Fisher, the Intel
software analysis group GM,

and also Michael Greene and a few others,
we bring the team together,
we started to look at

networking opportunities,
training opportunities.

We worked with our
industrial partners to offer

many free training classes and we also
start reaching out to universities
to build the pipeline.

And especially to motivate
the female students to get

passionate about big
data, about analytics.

So as of now, we have more
than 2000 members globally

for the forum and also
we have many chapters.

We have chapters along the West Coast
in the Bay Area, also East Coast.
We also have chapters in Europe and Asia
so we're definitely
seeing more and more women

getting excited with
big data and analytics.

And also, we have great collaboration
with women in data science at Stanford.
>> Yeah and it sounds like the momentum,
it doesn't sound like the
momentum, you can feel it, right?

You can feel it online with,
I can see a Twitter stream

in front of me on this monitor.
People are getting involved
in droves all across the globe

and I said to Margot, I asked her earlier,
Margot Gerritsen, one of the
founders of WiDS, I said,

first of all, you must be
pleasantly pretty shocked

at how quickly this has ascended.
And she said yes, and I said,
where do you go from here?

And she said, it's really
now going to be about getting

involved with WiDS more
frequently throughout the year.

Also, kind of going up a funnel
if you will, to high school

students and starting to
encourage them, excite them,

and start that motivation track,
if you will, even earlier.

And I think that is, in terms
to your point about we can't

do anything if the pipeline
isn't there to support it.

One of the things that
WiDS is aiming to do,

and it sounds like what
you're doing as well,

similar to Women in Big
Data Forum at Intel,

is let's start creating
a pipeline of women

that are educated in the technical side
and the software softer skill side
that are interested and find their passion
so that we can help motivate
them, that you can do this.

The sky's the limit where
data science is concerned.

>> Absolutely, absolutely.
And it's great to see
actually everybody recognize

the value of building the pipeline
and reaching out beyond
the university students.

Because have to get more and more girls
getting into the science and tech sector.
And we have to start from young.
And I, yeah, totally agree,
I think we really need to

build our pipeline and a
pipeline for our pipeline.

>> Yes, exactly.
And also that sort of sustaining momentum
as women, you know, go in university
and study STEM subjects,
get into the field.

Obviously retention is a big challenge
that the tech industry and
STEM fields alike have faced.

But that retention, that motivation,
and I think organizations like this,
just with this, you can feel
the passion when you walk

into this alumni center
at Stanford is really key.

We thank you so much for
carving out some time

to share your insights
and your career path

and your recommendations on theCUBE
and wish you continued success at Intel
and with Women in Big Data Forum,
which I'm sure we'll see
you back at WiDS next year.

>> Alright, thank you, thanks Lisa.
>> Absolutely, my pleasure.
We want to thank you, you have
been watching theCUBE live

from the Women in Data
Science Conference 2018.

Hashtag WiDS2018, join the
conversation, get involved.

I'm Lisa Martin from Stanford.
Stick around, I'll be right back with
John Furrier to do a wrap of the day.
(outro electronic music)


女性在資料科學 (Ziya Ma, Intel Corporation | WiDS 2018)

276 分類 收藏
alex 發佈於 2019 年 3 月 24 日
  1. 1. 單字查詢


  2. 2. 單句重複播放


  3. 3. 使用快速鍵


  4. 4. 關閉語言字幕


  5. 5. 內嵌播放器


  6. 6. 展開播放器


  1. 英文聽力測驗


  1. 點擊展開筆記本讓你看的更舒服

  1. UrbanDictionary 俚語字典整合查詢。一般字典查詢不到你滿意的解譯,不妨使用「俚語字典」,或許會讓你有滿意的答案喔