►
Description
We already know that software is eating the world. The real question has quickly become: how will companies use software to win? Hear from technologists from Ford Motor Company, GE Software, John Deere, and Target about how they are embracing software in unexpected parts of their companies. In a conversation moderated by GitHub Vice President of Product Kakul Srivastava, the panelists will discuss how their teams are using modern software-building tools to improve their businesses, why it's critical to build a world-class engineering team, and ultimately how to use software to deliver world class products and services.
Featuring:
Michael Davis, John Deere
Dragos Maciuca, Ford
Hima Mukkamala, GE
Samir Shah, Target
About GitHub Universe:
Great software is more than code. GitHub Universe serves as a showcase for how people work together to solve the hard problems of developing software.
For more information on GitHub Universe, check the website:
http://githubuniverse.com
A
B
Show
of
hands
how
many
of
you
would
like
put
GE
and
software
in
the
same
place.
Wow
half
of
them
are
g
people
yeah.
My
name
is
Hema
I
lead
the
platform.
Engineering
organization,
that's
based
out
of
san
ramon,
and
the
responsibility
of
this
team
is
to
build
a
platform
that
enables
customers
to
build
applications
that
connect
their
machines.
It
could
be
turbines
jet
engines
could
be
locomotives
and
sort
of
leverage.
The
value
from
that
data.
Before
that
I
used
to
work
at
sybase,
building,
mobile
platforms
and
application
servers,
I.
C
C
We
already
have
about
70
people
looking
at
getting
technologies
from
Silicon,
Valley
and
incorporating
them
in
vehicles.
Vehicles
become
more
and
more
software
driven
personally
I
started
my
career
in
Silicon
Valley,
so
I
I'm,
not
a
transplant
from
from
Detroit
before
Ford
I
work
for
apple.
Before
that
I
work
for
lockheed
martin
before
that
que
la
and
then
before
that
BMW
technology
office
one
day
were
first
opening
in
the
valley.
So
I
started
with
automotive
in
Silicon
Valley
and
now
I'm
back
to
automotive,
in
Silicon,
Valley,
I.
D
Am
Samir
Shah
I
manage
data
engineering
team
at
target,
with
focus
on
building
a
platform
for
predictive
analytics
one
and
building
a
product
on
top
of
it,
so
one
of
that
is
personalization
engine
that
we
built
recently
prior
to
that
I
was
at
demand.
Tech
company
was
focused
on
building
prising
life
cycle
engine
from
optimizations
perspective,
got
acquired
by
IBM
a
couple
of
years
back.
That's
about.
We
hi.
E
I'm
Michael
Davis
I
work
as
a
software
engineer
in
tech
lead
at
John
Deere
to
put
a
slightly
different
spin
on
the
first
question.
I'd:
ask
you
how
many
of
you
think
of
farmers
as
nerds
so
at
least
one
hand
fair
enough,
so
I'm
a
John
Deere
we've
been
delivering
this
with
technology
and
software
to
farmers
for
a
while
now
and
starting
in
nineteen,
the
end
of
the
90s
we
delivered
displays
to
farmers,
which
allowed
them
to
record
what
they're
planting
later
on
2002
self
driving
tractors.
E
We
have
GPS
that
will
get
accuracy
of
two
centimeters,
and
this
is
all
software
driven
stuff.
So
you
know
it
really
is
in
many
ways
a
software
company,
but
it's
growing
very
quickly.
So
my
history,
I
have
been
a
John
Deere
for
a
couple
of
years
now,
I've
worked
on
embedded
systems
inside
vehicles,
tractor
displays
and
now
I'm
working
on
mobile
apps
and
we
have
over
100
mobile
apps.
A
colleague
of
mine
just
told
me
available
now
on
android
and
iOS
and
yeah
before
that.
E
A
B
Announced
a
business
called
GE
digital
which
took
all
the
software
sort
of
bitter
scope
within
different
businesses
and
brought
them
together
as
a
software
company,
I
just
found
out
that
we
make
six
billion
dollars.
So
we
were
talking
about
it's
okay
to
invest
a
billion
dollars
if
we
can
make
a
six
billion
dollars
out
of
that.
So
it's
a
six
million
dollar
software
company,
hoping
to
get
to
top
10
software
companies
by
2020
roughly
translates
to
a
15
billion
dollar
revenue.
So
you
know
big
goals.
B
But
fundamentally
you
know
Jeff
ml1
said
it
was
like
couple
of
months
back
he's
the
CEO
of
GE.
We
went
to
bed
as
an
industrial
company
woke
up
as
a
software
and
an
analytics
company.
It's
a
profound
statement.
Every
industrial
company
in
the
world
is
starting
to
think
of
the
next
competitor
in
the
market
is
not
you
know,
a
different
industrial
company.
It's
the
software
players
that
are
going
to
eat
their
lunch
dinner
breakfast
everything,
and
so
we
all
had
to
become
a
software
company
to
retain
our
competitive
advantage
and
that's
the
goal
for
g.
C
Yeah,
so
the
shift
is
software
to
meet
it's
actually
very
personal
I
felt
it
on
my
own,
like
when,
when
I
was
a
BMW
in
two
thousand,
there
was
a
fundamental
shift
from
mechanical
engineering
to
electrical
engineering
in
in
the
automotive
industry.
Now
we're
going
through
this
only
15
years
later,
through
this
fundamental
change
from
electronics,
only
22
software
driven
so
people
keep
talking
about.
You
know
cars
being
computer
on
wheels.
It's
really
actually
more
like
network
on
networks
on
wheels
right.
C
We
have
easily
80
to
100
micro
processors,
they're,
all
interlinked
connected
and
and
now
we're
moving
towards
autonomous
vehicles
which
are
going
to
require
an
enormous
amount
of
software
to
process
all
the
information
from
lighters
to
radars,
two
cameras,
image,
processing,
sensory
version.
We
have
connected
vehicles
right,
you
have
all
the
data
that
goes
to
and
from
the
vehicle.
You
need
to
worry
about
security
in
you
to
worry
to
worry
about
privacy.
C
We
need
to
worry
about
data
analysis
right
all
these
vehicles
full
of
sensors
driving
around
they
can
collect
a
lot
of
information
about
the
environment,
about
the
cars
around
them.
To
make
the
experience
for
the
driver
much
better
than
it
is
today,
but
also
for
the
drivers
around
them.
For
for
us,
as
a
company,
the
user
experience,
we
can
improve
that
tremendously
again
through
software
and
and
being
able
to
to
change
on
on
a
yearly
basis
and
keep
improving
it.
D
Great
so
a
target
we
call
our
customer
guest
and
from
what
we
see
today,
guess
that
shops
digitally
and
on
stone
shops.
Three
times
more
than
just
people
who
stood
or
a
guest
who
Shops
at
store
only
with
94
million
Millennials,
who
are
digitally
connected
all
the
time
they
want
to
shop
from
wherever
they
want,
using
whatever
technology,
whether
it's
mobile
or
an
iPad,
or
even
on
your
machines
there
or
when
they
are
at
store.
They
are
still
connected.
D
They
demand
that
they
get
best
experience,
seamless
experience,
but
at
the
lower
costs,
and
that's
only
possible
by
doing
and
investing
in
a
technology
and
software
and
when
I
say
technology
includes
nor
in
software,
but
also
you
know,
Internet
of
Things
or
robotics.
So
these
are
things
which
are
going
to
drive
the
future
of
retail
space.
E
Yeah
so
John
Deere
I,
think
I
said
yesterday
is
a
company
that
sells
taxes,
we
don't
really
sell
software
and
you
know
that's
changing
and
in
many
ways
you
know
as
sure
the
same
applies
for
all
of
you
as
well,
but
a
company
like
john
deere
is
using
software
doll,
these
different
stages.
So
we
say
you
know
we
have
these
dealer
networks
and
we
provide
them
with
the
tools
to
sell
vehicles
and
we
have
lots
of
controllers
on
our
vehicles.
E
A
So
I
want
one
of
the
questions
I
had
for
use
as
you're
going
through
this
transformation
of
you
know.
Turning
these
companies,
these
totally,
you
know
different
industries
into
software
companies.
What
have
been
some
of
the
biggest
surprises
for
you
as
you
do
this
and
we'll
switch
this
around
Samir?
Why
don't
you
start
I.
D
Think
surprise
is
not
as
much
for
me
been,
you
know,
doing
software
for
a
long
time
been
related
to
soft
the
retail
industry,
but
argue
this
really.
The
focus
was
on
optimizing
supply
chain
or
at
the
enterprise
level,
how
you
do
merchandising
more
effectively
or
you
how
you
do
plan
effectively.
What
change
in
the
shift
has
happened
is
now
in
retail
is
really
guest.
Focus
guests
are
demanding
more,
they
want
things
now
and
you
know
as
a
lot
of
services.
D
If
we
make
you
know
if
they
are
able
to
get
the
goods
on
the
same
day,
probably
they
will
shop
sixty-four
percent
more
online.
So
really
it's
near
not
about
optimizing.
How
you
of
course,
sup
you
need
to
do
that,
we're
really
driven
by
guest
and
experiences
they
want
both
on
online,
as
well
as
our
in-store.
E
So
I've
been
a
John
Deere
for
two
years,
they're,
not
a
particularly
long
time.
One
of
the
things
that
has
surprised
me,
though,
is
the
techniques
for
management
and
process
control
that
have
been
used
very
successfully
for
manufacturing
and
design
of
vehicles.
Doesn't
they
don't
really
work
very
well
for
software
and
I?
A
E
I
can
tell
a
story
of
a
manager
of
ours.
He
come
to
help
manage
a
software
project
that
we
were
working
on
and
we
were
sitting
in
a
room
talking
about
how
this
new
process
was
going
to
work,
and
he
said
Owen
okay.
So,
when
this
team
is
finished
working
on
this
component,
they
can
move
on
and
do
something
else.
E
And
someone
asked
the
question
well
what
about
maintaining
this
and
he
said,
but
it's
done
once
it's
finished,
it's
finished
and
in
some
some
situations
this
might
be
the
case
if
you've
designed
a
piece
of
hardware
or
maybe
even
an
embedded
controller,
but
for
the
system
we
were
working
on,
which
was
a
embedded
linux
on
a
kind
of
7
inch
display
sitting
in
a
tractor.
It's
just
really
not
the
case
and
I
think
some
of
these
sort
of
the
idea
that
you
can
finish
a
software
project.
C
I
don't
know
if
I
call
it
surprise,
but
it's
to
us
it's
basically
the
two
sorry,
the
two
heads
we
need
to
keep
in
mind
is
the
one
being
in
the
automotive
industry,
which
has
to
provide
vehicles
that
are,
they
are
safe
and
regulated,
and
on
the
other
hand,
we
have
another
head
here
in
Silicon
Valley,
where
we
need
to
move
fast
so
but
I.
What
I
mean
is
for.
C
Exactly
so,
it's
one
thing
when,
when
you
have
you
know
your
your
favorite
software,
the
way
you
post
pictures-
and
it
goes
down
for
an
hour
any-
maybe
get
annoyed,
because
you
can
share
your
pictures.
It's
another
thing
when,
when
you
have
an
issue
with
your
car
and
it
stops
working
at
60
miles
an
hour
and
the
middle
of
the
highway
or
you
know
it's,
it's
a
completely
different
situation.
We
were
highly
regulated.
We
we
have
people's
safety
in
mind
for
us
station
of
such
thing
as
mini-me
Minimum,
Viable
Product
its.
C
C
B
Quick
story
just
want
to
join
G
when
I
came
back,
came
from
a
mobile
experience.
I
knew
everything
about
touch
screens
and
all
that,
so
we
had
to
build
an
application
for
factory
automation
and
so
the
store
the
story
leads
to
I
go
to
a
factory
that
does
locomotive
remanufacturing.
They
tear
apart
a
locomotive
all
the
way
down
to
its
bolts
and
then
the
application
you
are
trying
to
build,
used
to
be
used
by
this
technician:
who's
wearing
an
overall
wearing
gloves
and
touchdown
work.
B
B
We
talked
about
continuous
integration
deployment
and
all
that
here,
but
in
the
first
question,
I
get
asked,
is:
what's
your
20
year
roadmap,
the
the
Ville
engines,
with
the
20-year
roadmap
and
they're,
used
to
expecting
asking
those
questions,
and
so
we've
had
to
face
those
challenges,
educating
them
on
how
software
evolves
and
how
they
need
to
change,
and
we
have
done
some
good
work
with
Eric
raised
with
Lean
Startup,
bringing
down
the
cycle
of
how
they
build
their
assets,
industrial
equipment.
And
so
those
are
the
few
good
surprises.
A
A
You
know
one
of
the
things
that
we
spent
a
lot
of
time.
Thinking
about
at
github
is
a
developer
experience.
The
developers
really
core.
You
know.
Clearly
developers
are
important
parts
of
each
of
each
of
the
the
groups
that
you're
part
of,
but
your
developers
also
interact
with
this
broad
range
of
people
who
are
very
different,
whether
it's
mechanical
engineers
or
retail
staff.
A
B
So
one
of
the
biggest
things
that
we
had
to
do
was
educating
them,
like
I've,
said
on
how
they
should
look
at
their
requirements.
So
what
we
did
to
kick
it
off
when
we
started
was
we
bring
all
the
product
owners
from
these
businesses
who
gonna
consume
the
software
and
did
like
training
to
them
on
how
software
gets
built?
It
sounds
as
simple
as
that,
but
that's
the
reality.
B
If
you
come
from
an
industrial
engineering
background
or
a
rope
or
a
automation
background,
you
really
don't
know
what
a
July
agility
means,
and
so
the
lot
of
Education
had
to
be
done
collaboratively
because
they
are
an
integral
part
of
how
we
build
software.
They
give
us
a
requirements,
including
them
in
the
sort
of
the
whole
life
cycle.
It
was
very
important,
so
we
had
to
use
various
tools
to
enable
the
collaboration
so
that
they
play
an
integral
part
of
the
process.
C
Yeah
I
think
for
us
this.
We
have
the
same
same
challenge.
We
have
first
of
all
we're
a
global
companies
that
we
have
research
groups,
not
only
here
in
Palo
Alto,
but
we
have
them
in
dearborn
michigan
in
oven.
In
germany,
we
have
a
few
others
in
in
Asia
and
in
Australia,
so
collaborating
across
various
continents
and
requirements.
C
That's
one
of
the
challenges,
the
other
one
is
exactly
that:
the
difference
between
waterfall
and
agile,
so
most
of
the
stuff
that
we
did
in
in
in
the
building
cars
there's
been
waterfall
going
from
requirements
passing
them
on
to
to
the
tier
1
suppliers
trying
to
educate
people
in
doing
agile
development.
It's
been
probably
the
biggest
challenge
and
the
biggest
education
gap
that
we've
had
so
far,
but
with
tools
like
this,
it's
definitely
helping
quite
a
bit
to
bridge
the
gap.
D
When
story
is
no
different
in
any
other
industry,
but
what
I
see
the
biggest
shift,
that
is,
is
bringing
collaboration
between
in
our
case
merchants
and
engineers
and
making
them
as
part
of
one
team
and
providing
them
to
to
collaborate
and
bringing
them
in
a
platform
or
or
the
projects
where
they
are
part
of
the
team.
Earlier
it
was
very
much
like
what
he
mentioned.
We.
F
D
A
two
year
plan,
and
then
you
don't
see
they
give
you
requirement,
you
go,
create
a
plan
and
then
you
go
build
a
product
and
then
show
them
right
now
with
more
agile
processes
and
collaborations
they
are
seeing
the
progress
giving
you
feedback.
What
it
has
enabled
is
not
only
from
engineers
who
are
start
to
understand
domain
better,
but
also
what
it
has
done
is
from
the
business
owners
started,
understand
technology
better
and
now
it
has
opened
up
some
different
kind
of
interesting
idea.
D
A
E
So
I'm
to
agree
with
you
there.
The
idea
of
community
I
think
is
something
is
really
important
for
us,
because
these
people
are
coming
for
very
different
backgrounds.
You've
got
the
software
engineers
who
would
be
happy
writing
everything
in
markdown
and
then
other
people
who
are
used
to
using
word,
maybe
maybe
you're,
on
Microsoft
SharePoint,
whatever
it
is.
So
you
need
to
build
a
community
of
people
who
don't
have
a
common
tool
or
a
common
language
and
I
think
that's
really
tricky.
We
use
github
enterprise
and
its
really
helped
us
in
many
ways.
A
That's
great,
you
use
the
word
community
and
that's
also
a
word.
We
spend
a
lot
of
time
thinking
about
one
of
the
things
that
we've
we've
seen
as
as
different
industries
adopt
software
models
is
often
you
know.
You've
got
a
small
software
team
here
and
then
you've
got
another
small
software
team.
Here
you
know,
as
you
start
making
these
shifts
really
becoming
software
businesses
connecting
these
disparate
software
teams
becomes
really
important.
A
B
B
We
are
a
big
XP
shop,
extreme
programming
shop,
so
the
veil-
it
has
enabled
us,
when,
within
the
location,
to
sort
of
spread
the
knowledge
across
to
build
a
culture
where
the
teams
there's
not
one
focal
point,
one
person
who
knows
everything
about
a
topic,
so
we
do
promise
in
sparing
we
do
a
pretty
good
practice
to
distribute
that
knowledge,
and
so
that
has
helped.
You
know
to
to
build
the
teams
and
we
sort
of
follow
the
two
pizza
rule:
don't
grow
more
than
10
break
it
up.
B
I
mean
we're
not
perfect,
but
sort
of
approaching
these,
and
going
back
to
the
earlier
point
about
it's
funny
that
when
they
see
the
software
engineering
teams
practice
that
a
lot
of
the
hardware
engineers
know
want
to
practice
like
they
want
to
build
an
engine,
two
of
them
working
together.
So
if
they
got
excited
when
they
see
these
software
practices,.
C
Yeah
we've
seen
that
by
the
same
thing
here
to
its
we,
you
know
when
we
develop
apps,
we
develop
them.
Sometimes
we
develop
them
in
Palo
Alto.
We
showcase
them
and
the
prototype.
We
can
work
with
the
teams
back
in
Dearborn
to
to
actually
implement
and
scale
them.
We
work
with
tier
1
suppliers.
We
work
with
all
kinds
of
other
people
here
in
the
valley
to
to
help
us
with
different
aspects
of
that,
so
we're
trying
to
distribute
the
software
development
exactly
where
someone
has
a
core
competency.
Sometimes
we
do
different
changes
for
different
markets.
C
We
may
have
different
types
of
applications
for
let's
say
the
electrified
vehicles,
which
have
the
say,
different
information.
Then
then
you
would
have
for
a
gasoline-powered
engine,
so
you
know
for
the
gasoline
power
engine.
You
don't
need
to
know
when
you,
you
have
cheap
electricity
at
home.
If
you
have
a
plug-in
electric
vehicle
or
an
electric
vehicle,
then
you
do
need
to
know
that,
so
there
are
some
differences
depending
on
the
vehicle
models
depending
on
regions.
D
A
E
So
I'm,
unfortunately,
unlike
GE,
we
don't
have
a
single
co-located
site
where
all
of
our
teams
work
and
it
definitely
is
a
challenge
for
us
previous
product
I
worked
on.
We
had
maybe
200
people
in
the
state's
in
in
and
around
Des
Moines
and
50
people
in
Europe,
and
well
actually
also
a
couple
of
people,
maybe
60
in
India,
so
trying
to
bring
all
the
work
together
and
just
I
guess
the
integration
task
there
alone
was
really
difficult
and
the
communication.
E
We
we
started
to
use
campfire
and
then
we
were
told
we
weren't
allowed
to
use
that
and
then
there
was
a
different
tool
offering
that
we
got
which
we
were
allowed
to
use
and
we
started
we
jumped
on
there
and
that's
been
really
helpful.
So
this
sort
of
slack
style
messaging.
We
actually
use
flow
doc,
which
is
from
rally,
and
it's
been
really
really
helpful-
just
to
have
this
constant
stream
of
consciousness
from
a
development
team,
and
you
can
jump
in
and
you
say,
hey
guys,
you
know
our
prod
server
is
down.
E
What's
going
on
and
somebody's
there
all
the
time
to
answer
your
question
and
for
me
as
a
mobile,
app
developer,
now,
it's
great
because
I
go
in
and
I'm
saying:
hey
guys.
This
is
not
it's
not
running
and
somebody
from
the
platform
team.
Is
there
answering
my
questions
addressing
my
issues
immediately,
whereas
before
it
might
have
been,
I'd
have
to
go
somewhere.
E
A
E
E
Bringing
github
to
john
deere
was
an
interesting
sort
of
story.
We
officially
used
subversion,
so
we
had
a
multi-site
subversion,
install
lots
of
licenses
and
it
was
the
standard
for
the
organization.
I
think
it
became
the
standard
in
2012,
so
slightly
behind
the
time
and
the
project
I
was
working
on
actually
used.
Mercury,
oh
and
that's
much
better
than
subversion.
E
You
know
we
had
50
60
70
users
from
across
the
organization
using
it
and
suddenly
it's
too
late
to
go
back
officially
the
company
policy
would
have
been
no
don't
do
that
and
so
I
think
that's
one
of
the
situations
where,
by
giving
somebody
permission
to
go
ahead
and
just
try
something,
you
really
get
a
huge
benefit,
and
now
we
have
I
think
about
it.
Bounced
1300
seats
or
get
up
enterprise
across
different
servers
and
people
don't
want
to
go
back.
So
we.
A
Like
to
hear
that
cool
I
want
to
open
up
for
questions
in
the
audience,
one
ask
one
more
so
start
thinking
about
your
questions
and
we'll
get
we'll
get
going
to
that.
There's
this
idea
of
giving
permission
and
seeing
what
comes
up
you
know
from
from
the
team's
themselves,
in
terms
of
serendipitous
innovation,
you
know
building
these
communities
across
the
company.
You
know
having
these,
you
know,
Adams
and
neurons
sort
of
interacting
with
each
other.
B
So
one
of
the
things
that
you
know
it's
very
hard
as
a
leader
to
let
go
to
lose
control,
and
so
that's
what
we
struggle
with,
especially
coming
from
GE,
which
is
a
top-down
sort
of
command
control
management.
But
one
of
the
biggest
changes
we've
seen
is:
we
went
we
sort
of
pivoted
to
a
micro
services
architecture
and
we
have
eight
teams
in
the
platform
space
using
eight
different
languages
to
build
those
services.
This
could
have
never
happened.
Going
back
to
the
John
Deere
example.
B
This
would
have
not
happened
two
years
back
right,
so
the
developers
have
the
power
to
choose
the
language,
to
choose
the
platform
to
choose
the
database
that
best
suits
whatever
capabilities
they
provide
and
so
that
transformation
of
organization-
and
we
also
look
at
that
as
the
biggest
challenge
the
Millennials
get
it.
But
we
have
a
broad
mix
of
sort
of
engineering
teams
where
they
are
used.
B
If
you've
been
in
this
industry
long
enough,
you're
told
to
do
how
you
know
there
isn't
a
notion
of
team
empowerment,
team
power,
team
decision
making
and
so
bringing
that
culture
is
harder
than
what
people
think
it
is,
and
so
we
spend
a
lot
of
time
telling
the
teams
it's
okay,
to
fail.
It's
okay,
to
make
those
decisions,
leave
us
out
of
the
leadership
out
of
the
way
and
the
teams
get
to
make
those
decisions
and
that's
the
transition
that
we
are
in
transformation
we
are
in,
which
will
take
time.
Yeah
yeah,.
C
For
us,
it's
very
similar,
so
special
hearing
in
Palo
Alto.
We
have
also
the
curse
of
the
distance
because
we're
removed
away
from
the
for
the
motherships
of
communication
it's
hard,
but
at
the
same
time
we
a
lot
more
freedom
there.
We
don't
have
people
looking
over
our
shoulders
telling
us
what
not
to
do
so.
We
have
several
teams
that
the
same
thing
they're,
using
whatever
tools
they
want.
They're
developing
gaps,
are
developing
new
hardware,
they're
developing
mechatronics
and
then,
when
something
cool
comes
out,
we
just
showcase
it.
C
My
favorite
story
is
how
we
we
ended
up
with
the
Apple
watch
app
for
for
Ford,
because
I
had
a
guy
that
was
fresh
out
of
school.
He
just
came
one
day
as
he's
like
this
was
even
before
Apple
released
a
watch,
so
I
can
I
develop
an
app
for
it.
I'm
like
sure,
go
ahead,
and
then
two
weeks
later
he
had
a
prototype.
We
had
a
visit
from
the
CEO
and
he
ended
up
playing
with
a
watch
for
at
least
half
an
hour.
C
He
almost
stole
the
watch
from
him
and
you
know
then
we
pushed
it
out
to
to
dearborn
for
for
full
development,
and
now
you
can
get
it
from
my
tunes.
So
it
was
this
this
kind
of
empowerment
and
freedom
that
allowed
us
to
really
quickly
develop
an
app
that
you
know.
People
can
now
have
and
start
their
cars
and
lock
doors
and
do
all
kinds
of
things
with
it.
That.
D
Me
it
is
really
changing
a
mindset
from
my
to
be
on
I
think
a
lot
of
those
you
know.
I
can't
give
credit
to
engineers,
and
you
know
who
went
ahead,
build
something
and
then
showcasing
it.
It
can
be
done
and
you
know
get
enterprise
that
what
we
use
really
was
got
started
by.
One
person
then,
was
in
the
audience
today
and
he's
going
to
talk
tomorrow
about
that.
D
Well,
he
went
ahead
and
brought
it
in
and
started
showcasing,
and
now
we
have
three
thousand
software
engineers
building
and
committing
coding
to
get
so
I
think
that's
what
it
is
empowering
engineers
to
go:
try
out
new
things,
embracing
open-source,
more
bring
creativity
and
let
them
come
back
and
show
what
can
be
done.
I
think
that's
a
key.
E
A
F
Alright,
so
Chris
this
morning
talked
about
software
being
kind
of
a
new
frontier
or
the
next
frontier,
and
so
this
is
really
for
the
hardware
guys
we're.
In
recent
years.
We've
we've
made
a
lot
of
advances
in
how
we
develop
software.
We
have,
we
have
essentially
the
ability
to
escape
a
lot
of
physical
limitations.
We
can
scale
up,
we
can
scale
out.
F
We
can
scale
time
I'm
curious,
how
the
hardware
guys
are
trying
to
bring
some
of
those
concepts
to
the
hardware
they
work
on,
which
is
really
constrained
by
the
physics
and
the
reality
of
things.
Are
there
any
interesting
challenges
or
techniques
that
you
guys
have
employed
to
to
kind
of
escape
those
boundaries
like
3d,
printing
or
any
other
kind
of
ideas?
So.
B
I,
give
you
my
perspective
and
in
our
sense
our
hardware
is
a
locomotive.
So
in
that
context
you
know
when
we
run
the
analytics
the
data
science,
the
whole
value
proposition
that
we
want
to
provide
is
but
get
a
example
of.
How
does
the
turbine
react
to
wind
and
pressure
right?
So
you
want
to
take
that
feedback.
Give
it
back
to
the
design
people
give
it
back
to
the
physics,
people
and
sort
of
make
new
composites
that
react
better.
B
C
Yeah
for
us,
it's
mostly
simulation
right,
so
you
can
build
an
entire
car
crash.
It
spin
it
all
around
cut
it
in
pieces,
and
it's
all
done
in
software
before
you
even
play
with
the
first
piece
of
piece
of
metal,
so
the
number
of
iterations
you
can
happen
that
the
speeder
you
can
you
can
design
new
new
cars
through
software
simulation.
It's
it's
amazing.
C
Now
we're
good,
sometimes
we're
running
the
risk
of
of
just
keeping
it
in
the
computer
too
long
before
we
actually
release
it
in
hardware,
but
the
amount
of
simulation
that
we
can
run
and
a
fidelity.
Do
you
get
out
of
it?
It's
it's
simply
amazing
and
that's
one
of
the
ways
it
you
know
it
allowed
us
now
to
have
the
f-150.
That's
all
aluminum
body
that
nobody
thought
you
can
do
it
and
no
new
materials,
new
processes.
F
G
So
it
was
mentioned
actually
that
that
you
can't
build
a
car
through
an
MVP.
You
can
build
a
software
product
for
a
car
through
an
MVP.
Well,
can
you
elaborate
a
little
bit
more
on
that
because,
like
what
did
it,
all
mask
did
with
Tesla
what
what
is
Google
doing
right
now
with
Google
car
ain't,
that
and
I'm
VT,
and
how
is
the
auto
industry
going
to
learn
from
the
software
industry
in
that
respect,
I'd
make
the
other
otherwise
other
way
around.
C
Again,
I
mean
typically
when
I
talk
about
cars,
and
you
talk
about
ten
VP,
it's
what
the
Model
T
right.
I
mean
we've
done
that
a
hundred
years
ago,
so
you
can't
go
back
to
that.
So
I
wouldn't
say
that
that
does
like
released
an
MVP
right
it.
It's
a
real
car
that
needs
to
meet
crash
standards
and
it
needs
to
meet
all
kinds
of
other
regulations
and
it
needs
to
be
safe.
So
you
can't
go
to
you
know
to
the
public
and
test
all
kinds
of
crazy
ideas
you
need
to
have.
C
You
need
to
come
up
with
something
that
that's
already
solid.
That's
already
been
tested
that
already
meets
a
lot
of
requirements
that
in
software
you
can
do
the
the
other
way.
But
you
know
we
can
always
try
the
other
things.
I
mean
I'm,
thinking,
OAB
testing.
We
can
probably
start
doing
some
of
that,
especially
when
it
comes
to
to
the
infotainment
system.
You
can
do
a
be
testing
on
I
know
the
ABS
system
cuz.
If
one
doesn't
work
in
it,
it's
not
acceptable.
C
But
in
terms
of
you
know,
what
do
you
arrange
the
buttons
in
theory?
We
could
we
could
do
that,
then
again,
there
are
implications
that
that
does
a
shoe
it
with
automotive.
Is
that
even
the
buttons
heavenly
like
software
buttons
I
mean
on
the
Ivy
I
have
implications
in
terms
of
safety.
If
you
can't
find
the
right
button,
you
stop
fiddling,
and
you
spend
more
than
two
seconds
trying
to
change
the
radio
station
that
in
itself
it's
not
just
annoying.
It
becomes
a
safety
you.
C
F
B
Can
I
get
your
email
address
so
the
way
we
look
at
this
is
ameena
industrial
companies
and
I'm,
including
john
deere,
GE
and
others
in
this
mix,
is
what
you
get
to
do
is
pretty
interesting
working
at
a
company
like
this
right,
yeah
we're
getting
to
work
with
and
all
our
challenge.
Fascinations
include
trains
and
planes
and
you're
working
with
software.
That's
dealing
with
that
its
life.
You
know
it's
life,
changing
that's
how
we
try
to
compete
with
them.
Where
are
the
traditional
social
and
other
companies
you're
building
stuff?
B
C
No,
I
mean
for
us
it's,
it
is
hard
but
like,
like
I
said
what
what
we
offer
in
exchange
is
special
for
those
who
are
passionate,
but
cars
right.
You
can
actually
drive
and
and
see
the
things
that
you
worked
on
and
it's
kind
of
cool
to
change.
Does
it
that
the
software
and
then
seeing
it
implemented
in
an
actual
vehicle
they
can
drive
and
show
it
to
your
friends
what
what
you
did
so
so
that
passion
for
for
the
final
product
I
think
it's
what
differentiates
us.
D
D
Think
it's
really
a
problem
that
you
are
trying
to
solve
and
magnitude
of
impact
that
you
can
make
to
the
business
and
once
you
are
able
to
kind
of
show
the
value
and
the
work
that
they
will
be
doing,
I
think
a
lot
of
people
can
relate
and
say
my
work
can
make
such
a
huge
difference,
whether
you
know
making
a
better
community
or
you
know
better
business
I
think
that's
where
you
would
see
we're
defining
your
problem
better
and
making
it
purposeful
and
I.
Think
that's
what
attracts
a
good
talent.
E
I
think,
is
a
different
sort
of
there's
a
different
value
proposition
there
and
lots
of
the
people
who
work
for
us
and
who
are
interested.
They
may
be
a
have
been
farmers
in
the
past
or
their
family
was
in
farming
or
you
know
they
really
like
driving
around
tractors.
Cuz.
That's
pretty
much!
That's
a
good
time
right!
No
just
you
can
sit
in
one
of
these
things
and
it's
like
40
tons
is
a
couple
of
hundred
horsepower
and
you
just
know
you
could
drive
through
anything
and
yeah.
A
I
What
experiences
have
you
had
dealing
with
communication
challenges
across
disciplines
and
then
what
advice
would
you
have
for
someone
like
me
who's
one
of
two
developers
at
a
200-person
behavioral
science
company,
and
we
have
challenges,
communicating
value
of
features
that
we
put
in
software
and
why
it
might
be
valuable
for
us.
The
right
might
be
valued
by
valuable
to
the
scientists
or
data
analysts.
I
E
Experience
has
become
a
really
big
thing,
and
so
the
idea
that
hey
you
know
your
vehicle,
your
tractor,
it
can
be
friendly
to
you
know
it.
Can
you
can
look
at
it
and
think?
Well,
maybe
if
I
push
this
button
is
just
going
to
work
and
there's
obviously
the
safety
aspect
of
that
it's
got
to
be
very
clear
and
easy
to
use
and
it
can't
be
ambiguous,
and
so
that's
one
place
where
we've
really
pushed
out.
E
B
One
of
the
things
that
has
helped
does
is
we
always
look
at
it
from
what
value
does
it
provide
to
the
customers?
So
the
way
we
look
at
it
is
our
whole
reason
of
investing
in
software
is
to
enable
unclad
reduce,
unplanned
downtime.
So
every
question
that
we
ask
visas,
the
requirement
is
driven
by
how
much
or
will
it
help
customers
and
run
their
planes
on
time,
run
their
trains
on
time
and
that's
what
helps
us
prioritize?
The
second
thing
which
I
want
to
live
with
this
transparency.
B
You
know
these
organizations
have
used
to
having
control
over
everything
so
breaking
down.
Those
barriers
has
been
very
important
so
that
we
completely
expose
what
our
backlog
is,
how
we
prioritize
they
attend
our
sprint
planning
that
and
our
release
planning
so
making
it
a
very
transparent
process,
helped
them
realize
how
they
play
a
role
in
it.
A
C
Don't
think
I'm
really
excited
about
what
we
have
here
in
Palo
Alto
is
that
we
don't
have
only
engineers.
We
have
maybe
sixty
percent
engineers,
but
we
also
have
user
experience.
We
have
industrial
design,
we
have
ethnographers,
we
have
marketing
and
sales,
so
we're
all
working
together
to
develop
these
products.
It
again
from
from
a
very
personal
experience
when
15
years
ago,
when
I
was
a
BMW
and
we're
only
a
bunch
of
Engineers.
What
do
engineers
do
that?
C
I've,
love
to
throw
features
right
feature
feature
feature
without
being
coherent
as
soon
as
we
got
one
marketing
person
that
finally
guided
us,
it's
like
well,
this
makes
sense
to
me
do
this:
what's
all
this
will
not
sell
this
meets
the
brand
is,
doesn't
need
the
brand
so
be
drinking
all
this
sort
of
softer
side
of
the
business
to
develop
software
together
with
with
with
the
engineers
I
think
it's
that's
a
winning
proposition
because
you're
you
otherwise,
as
I
said
engineers
love
to
do
features
that
may
now
make
sense
for
the
final
user.
I.
D
D
There
is
no
human
emotion,
really
data
speaks
and,
if
you're
able
to
prove
that
it's
either
you
really
building
a
product,
that's
going
to
make
a
difference
or
or
not,
and
whether
you
want
to
continue
or
not
and
engage
right,
engage
your
customers
where
you
know,
if
you
are
doing
something
going,
a
new
field
which
is
going
to
be
implemented,
say
we
have
a
sign
and
you
want
to
apply
that
to
marketing.
Have
them
come
and
tell
what
they're
trying
to
solve?