►
From YouTube: INDUS Driving Digital Transformation - Google's financial transformation using SAP Central Finance
Description
In this session, attendees will have a chance to discover how Google transitioned from a non-SAP ERP to SAP S/4HANA and discover some lessons learned on:
Leveraging SAP Central Finance to migrate from Oracle to Greenfield SAP S/4HANA
Deploying SAP S/4HANA Cloud for subsidiaries
Making the most of the Google Cloud Platform’s hyper-scaling nature
Name and a short bio of the speaker
Julien Delvat is a Solution Architect at Google where he contributes to the Enterprise Transformation project. From developer at SAP Labs France to SAP consultant around the globe and for various industries, Julien shares his 20 years of experience in the SAP world, especially around S/4HANA, S/4HANA Cloud, and Central Finance.
A
Years
was
what
I
spent
with
sap
in
singapore,
covering
you
know,
covering
and
talking
with
the
customers
in
the
southeast
asia
region
doing
various
roads.
You
know
in
the
in
the
pre-sales
space,
and
I
look
forward
to
you
know
connecting
connecting
with
all
of
you
after
after
this
session,
so
over
to
you,
rohan.
B
Yeah
thanks
so
good
morning,
good
morning
and
good
afternoon,
everyone
I'm
rohan
desai.
So
currently,
I'm
leading
sap
technology
for
google
cloud
india
and
we
I
myself
and
naren,
are
part
of
the
same
team
so
because
so
this
is
also
my
18
months
in
google
cloud
focusing
on
sap
and
before
joining
google.
I
was
part
of
sap
india,
so
for
for.
C
C
You
yeah,
thank
you.
I
hope
you
can
hear
me
very
well,
I'm
not
sure
if
my
camera
is
going
to
be
happy
with
me,
but
that
doesn't
matter
as
much
so
I
am
everybody.
My
name
is
julian
dama.
I
know
mukesh
said
it's
it's
hard
to
get
to
know
people,
but
I
really
appreciate
the
fact
that
I
can
you
know
I
don't
have
to
jump
all
the
way
to
india
or
singapore
to
do
this
session
for
an
hour
and
a
half
and
then
go
back
home.
C
So
I'm
calling
you
here
from
san
francisco,
hopefully,
hopefully
the
the
time
is
working
for
all
of
you.
I
have
been
with
google
for
two
and
a
half
years
now
working
on
our
finance
transformation
project
and
I
will
go
into
more
details
here.
I
actually
have
20
years
experience
in
the
sap
world.
I
started
at
sap
in
france
as
a
developer
in
finance
and
then
went
into
consulting
based
out
of
the
u.s
but
went
all
around
the
world
in
multiple
industries.
Multiple
countries,
including
india,
not
yet
singapore.
C
A
C
Like
google,
I'm
pretty
sure
that
your
companies
are
more
than
one
product
the
solution
we
had
to
come
up
with
had
to
cover
not
only
google
but
the
whole
alphabet.
Spectrum,
which
includes
companies
that
do
have
a
lot
of
revenue
like
you
know,
youtube
or
nest
the
thermostats,
and
also
companies
that
are
still
pre-revenue
or
companies
of
all
sizes,
like
waymo,
being
a
self-driving
car
company
being
obviously
much
smaller.
C
C
C
Our
goal
was
to
go
fast
to
market,
and
so
what
happened
was
that
the
infrastructure
grew
with
leaps
and
bounds,
with
a
very
large
number
of
systems
that
were
not
always
connected
together.
So
it
was
always
hard
to
reconcile
you
know:
company
codes
and
cost
centers
and
and
geo
accounts,
and
and
how
different
parts
of
the
organization
were
tracing
different.
You
know
elements
in
the
finance
area,
so
we
needed
to
do
something
because
what
what
got
us?
C
Call
for
transformation
to
reduce
the
friction,
to
really
improve
the
the
return
on
effort
right,
get
more
output
out
of
our
people
being
easier
to
manage
and
to
reduce
the
risk
and
still
keep
excellence
at
everything
we're
doing
right.
So
I
think
this
is
a
picture
that
you
probably
see
in
a
lot
of
areas
and
hopefully
not
as
bad
in
your
organizations,
but
it's
not
that
exceptional.
C
She
gave
us
a
three-year
objective
that,
within
three
years,
we
will
have
even
more
useful
and
accurate
data
at
our
fingertips
and
the
processes
and
technology
to
maximize
business,
velocity
and
growth,
and
if
we
break
down
this
this
sentence
here,
what
you
find
here
is
all
around
data.
This
is
extremely
important
at
google.
This
is
extremely
important
at
sap,
but
you
have
to
make
it
available
and
useful.
This
is
actually
google's
mission
to
make
the
world
information
available
and
useful
and
how
we
do.
C
E
C
Was
kind
of
a
a
challenge
right,
the
vision
is
really
to
enable
our
finance
googlers
femalers
to
deliver
their
best,
so
to
enable
them
their
systems
around
them
were
restricting
them,
and
to
do
that,
we
need
to
create
an
infrastructure
to
support
alphabet
that
200k
at
200
billion
dollars.
What's
interesting
is
that
message
was
given
to
us
in
2019
when
we
were
80
billion
dollars
and
we
are
at
160
now
so
we're
pretty
close
to
actually
the
200
billion
now
in
terms
of
systems.
C
E
C
Back
together
all
the
time
we
wanted
to
get
out
of
the
box
so
back
to
standard,
so
we
don't
have
to
do
custom
development
because
you
know
what
makes
google
google
is
is
new
products
on
the
market.
It's
not
about
how
we
find
taxes
right.
So,
let's
get
out
of
the
box
as
much
as
possible
and
then
fewer
systems,
fewer
moving
parts,
basically
reducing
complexity
of
the
systems.
B
C
Was
the
scope?
What
did
we
have
to
do?
Well,
it
was
really
about
everything
right.
So
product
areas
are
like,
like
divisions
in
the
organization
right
so
way,
more
verily
had
very
special
requirements.
So
how
do
we
treat
them
differently
start
with
our
treasury,
which
was
one
of
the
largest
asset
management,
was
a
project
that
we
had
prepared
before
our
hardware.
Our
device
and
services
organization.
Has,
you
know,
inventory
requirements
they
have
revenue
requirement
or
revenue
concerns.
So
how
do
we
do
that
part
of
the
world?
C
This
is
very
different
than
the
technical
infrastructure,
which
is
a
lot
of
procurement
and
assets
and
building
the
infrastructure.
So
you
know
every
one
of
those
have
very
different
priorities
and
different
ways
to
to
solve
them,
but
still
we
want
them
to
be
integrated.
So
how
do
we,
you
know
mitigate
the
two?
C
Our
finance
score
covers
the
whole.
You
know
aspect
of
finance,
standard,
gl,
inter
company
cash
tax
fp,
a
international
for
us
was
very,
very
important
because
our
previous
erp
system
wasn't
really
delivering
on
that
aspect.
There
was
a
whole
shadow
system
going
on
and
fixed
assets,
and
cost
inventory
was
very
important
because
of
the
volume
very
simply
so
to
pay
the
whole
procurement
part
had
to
be
reinvented.
C
Basically,
the
idea
was
to
be
able
to
start
having
you
know:
cash
discounts
having
more
buying
power
in
terms
of
you
know,
collecting
and
grouping
orders
and
international
payroll
in
the
us.
We
recovered
us
in
canada,
but
in
terms
of
international
payroll,
we
wanted
to
rethink
the
way
we
were
doing
it
in
18
countries.
Those
are
like
the
major
pillars
of
our
project.
C
They
all
relied
on
some
horizontal
work
streams,
change
management,
that's
nice
to
mention
this
is
a
very
important
topic
for
us
controls,
of
course,
together
with
security
technology
which
is
really
about.
How
do
we
enable
connect,
those
systems,
data
and
reporting?
And,
of
course,
international
operations
was
also
horizontal,
because
it
has
an
impact
on
all
the
other
killers,
all
right
timeline
we
started
from
nothing.
I
think.
That's
that's.
The
key
message
here
is
that
when
we
started
this
project
we
had
zero
sap
right.
C
It
was
all
in
in
another
system,
so
we
had
to
start
somewhere
and
maybe
you
know
in
a
weird
way.
We
started
this
treasury
to
get
our
fit
in
the
water.
If
you
will,
the
treasury
plus
mdg
were
the
first.
We
started
to
put
together
to
have
something
we
can
control
that's
manageable
in
terms
of
size
to
get
to
know
sap
on
gcp,
because
we
had,
you
know
no
sap
basis,
people.
We
had
no
sap
developers
or
very
little
so
like.
Where
do
you
start?
C
You
start
with
something
small,
and
then
you
expand
that's
what
we
did
in
2019.
Then
we
started
with
our
finance
finance.
The
goal
was
to
really
change
the
system
of
record
by
our
finco
release.
2
and
we
had
a
technical
go,
live
with
finco
release,
one
I'll
go
back
to
that
a
little
later,
but
you
can
see
that
we
had.
C
C
Let's
see
now,
let's
take
the
example
of
treasury
to
see
you
know
what
that
means.
What
are
the
volumes?
What
are
we
talking
about
here
so
so?
Treasury
at
google
is
complex,
mostly
because
of
size,
and
we
wanted
to
automate
and
simplify
all
the
processes
and
transactions
in
a
controlled
environment,
because
you
don't
really
want
to
mess
with
cash
right.
C
So
we
have,
you
know
almost
100
legal
entity.
We
had
you
know
dozens
of
bank
accounts,
90
billion
dollars
us
dollars
in
cash,
hundreds
of
billion
dollars
in
national
functional
currency,
hundreds
of
billions
of
cash
flow
to
handle
in
a
project,
so
massive
amounts
massive
scale,
and
so
how
do
we
do?
And
how
did
we
do
it?
The
results
here
was
really
the
the
goal
was
to
eliminate
a
key
person
risk.
You
know,
how
do
you
reduce
the
risk?
C
C
Also
try
to
instead
of
going
to
each
bank
portal
trying
to
have
a
common
framework
for
all
of
them,
then
do
end-to-end
processes
right
so
all
on
and
on
and
on
until
we
automate
our
cash
position
and
improve
our
forecasting
by
legal
entity.
That
was
really
the
goal,
and
then
we
have
now
a
more
scalable
operating
model
so
basically
go
from
a
very
manual
very
spreadsheet-oriented
process,
very
high
risk,
very
high
effort
to
an
automated
transactions
and
reporting,
and
that's
scalable
for
the
future.
C
C
You
define
the
stakeholders,
you
define
the
risk
and
then
you
go
about
what
are
the
options,
considerations
pros
and
cons,
and
then
you
come
up
with
a
recommendation
that
you
bring
to
a
board
when
you
do
that.
What
happens
is
that
a
subgroup
can
take
care
of
it
and
then
bring
it
back
to
the
larger
group
for
communication
so,
and
what
also
is
interesting
is
if
somebody
comes
in
the
project
later
on,
they
can
just
go
back
to
the
decisions
and
see.
C
Yes,
we
considered
all
the
alternatives,
and
this
is
why
we
didn't
go
with
option
a
and
option
b
instead,
so
we
had
very
different
types
of
of
decisions
in
terms
of
the
eye.
You
know:
what
do
we
do
with
our
legacy
system?
Do
we
put
everything
in
bw?
Do
we
put
it
somewhere
else
and
I'll
come
back
to
that
boundary
systems
like
what's
the
right
level
of
granularity?
C
How
do
we
integrate
them
in
terms
of
consolidations?
You
know:
what's
the
master
data,
what's
the
model
and
a
number
of
decisions
here
on
coregl
around
the
master
data
definition
that
was
very
important,
obviously
around
fp,
a
like:
what's
the
allocation
strategy
which
tool
are
we
using
and
then,
if,
once
we
selected
the
two
like
pipm,
how
do
we
install
it?
C
C
It
was
a
lot
easier
to
keep
them
on
their
own,
separate
erp
system
and
then
integrate
them
to
the
main
one.
So
something
to
consider,
of
course,
the
cloud
solutions
like
aribao
anatix
cloud
and
concur
all
connected
to
our
main
system,
and
then
we
have
a
bwbc
system.
What
you
see
in
gray
are
the
things
that
we're
going
to
have
in
the
future
like
never
near
revenue,
accounting
or
advanced
blank
reporting,
for
instance.
C
I
frequently
have
the
question:
why
bpc
and
not
not
group
reporting?
Very
simply
when
we
started
the
project
we
felt
like
group
reporting
was
not
cooked
enough,
so
if
bpc
was
here
was
robust
was
reliable.
So
that's
that's
what
we
went
with
and
and
in
the
future
we
might
revisit
the
decision.
But
for
the
moment
this
is
this
is
what
we
have.
E
C
C
We
connected
an
slt
system
to
an
s4hana
box
and
we
also
had
the
vw
bpc,
because
really
the
core
of
our
minimum
valuable
product,
if
you
will,
was
the
ability
to
perform
consolidation
over
two
quarters,
that's
what
we
had
to
do.
We
also
started
to
establish
our
bi
framework
around
the
gcp,
bigquery
and
I'll,
come
back
to
the
benefits
of
that
to
connect
to
our
boundary
systems.
C
So
that
was
the
initial
system
when
we
went
live
with
the
first
release
phase,
one
in
february
of
2020.,
everybody
was
in
oracle
and
all
the
transactions
were
replicating
in
all
of
the
downstream
systems,
and
we
did
that
for
two
quarters,
because
we
had
a
lot
of
changes
in
the
in
the
enterprise
structure,
a
lot
of
changes
in
the
processes.
So
we
wanted
to
make
sure
that
everything
was
running
smoothly
and
we
did
that
for
two
quarters.
C
That
went
very
nicely,
so
we
changed
the
system
of
record
from
article
into
sap
in
july
of
2020,
and
this
is
when
the
core
finance
team
moved
out
of
oracle
into
sap,
which
means
that
all
inter-company
transactions,
for
instance,
were
done
here
right
and
now
we
started,
then
the
phase
three,
which
is
all
what's
called
the
sub
ledgers
in
oracle.
Basically,
the
migration
of
assets,
inventory
apar
and
everything
that
was
still
performing
in
oracle.
C
C
Realistically,
we
might
keep
the
slt
system
because
we
might
want
to
integrate
with
other
systems.
For
instance,
we
just
acquired
fitbit,
so
the
question
would
be:
do
we
integrate
with
bpc
for
only
a
consideration
level?
Do
we
integrate
at
an
s4
level,
or
do
we
integrate
an
slt
level?
Now
we
have
options
right.
So
that's
something
we
might
keep
in
our
back
pocket,
but
at
least
we
can.
You
know
very,
I
mean
highly
reduce
the
the
sizing
of
that
system
and
then
keep
it
quiet
right.
C
C
One
of
the
key
things
that
we
had
to
solve,
though,
was
the
mapping
right
because
we
go
from
an
old
world
to
a
new
world
in
the
old
world.
We
talk
about
legal
entities
in
the
new
world
of
sap.
We
talk
about
company
codes,
you
go
from
shadow
of
account
to
really
a
combination
of
profit,
central
cost,
centered
gl
accounts.
You
go
from
department
to
call
center.
So
really
this
whole
language
dependency
had
to
be
resolved.
C
There
are
a
number
of
ways
to
do
it.
Slc
actually
has
one
way
to
do
it,
but
in
our
case
what
we
wanted
is
to
have
many
too
many
systems
and
in
both
ways.
So
what
we
did
is
that
we
created
our
own
mapper,
which
is
really
a
custom
application
on
top
of
gcp
tools,
where
we
could
send
data
either
through
an
api
or
through
flat
files
and
then
get
them
translated
from
left
to
right
and
from
right
to
left.
C
C
Slt
is
one
great
option
to
do
it,
but
it's
really
mostly
one
way
from
old
to
new
and
then
from
either
sap
and
sap
to
sap
going
back
can
be
a
challenge.
So
in
our
case
we
had
developed
this.
How
to
do
that.
Basically,
you
can
rely
on
the
slt
system.
The
slt
system
from
sap
has
staging
tables
for
external
systems,
so
cf
stands
for
central
finance
and
e
for
external
right,
so
you
can
connect
to
that.
C
Is
push
the
the
rules
into
the
data
transformation
system,
so
something
to
consider
if,
if
you
are
considering
central
finance
and
slt,
something
that's
also
out
of
the
ordinary,
I
don't
know
if
you
heard
about
this
solution
called
papm
from
sap
profitability
and
performance
management.
I
think
I
always
mix
up
the
letters
here.
So
don't
quote
me
on
it,
but
I
think
it's
profitability
in
performance
management.
We
use
it
extensively
for
internal
company
transfer,
pricing
and
allocations.
C
The
idea
is
that
this
is
a
pw
based
application,
that
is
a
location
engine,
so
we
get
the
data
from
from
sap
for
the
actuals.
We
get
it
from
other
boundary
systems
for
our
plan
and
we
get
drivers
into
the
vw
system
and
we
can
crunch
the
number
and
then
you
get
the
data
back.
This
in
our
case
was
much
much
more
performant
than
all
the
other
solutions
that
we
have
evaluated.
C
So
if
you
have
never
heard
of
it,
I
highly
recommend
you
take
a
look
all
right
now,
something
that
is
very
interesting
for
us.
But,
as
I
said,
we
wanted
to
go
back
to
standard,
especially
for
the
things
that
you
know.
We're
not
unique
right
so
booking
an
inter-company
transaction
is
is
not
something
that
will
make
us
very
different.
However,
our
billing
system
is
something
that
we
believe
is
very
important
to
us
right,
especially
because
we
can
deal
with
micro
payments
at
scale
like
nobody
else.
C
So
this
green
box
here
the
billing
system
is
our
custom
solution.
We
didn't
want
to
touch
it.
The
question
is:
what
do
you
do
with
that
data,
because
you
need
some
of
it
into
your
s4
system
now.
Do
you
bring
everything
or
do
you
bring
a
piece
of
it
and
if
you
do
bring
everything,
that's
gonna
cost
a
lot
of
money,
and
if
you
only
bring
pieces
of
it,
then
how
do
you
reconcile
it
so
that
you
can
do
a
pnl?
C
So
now
we
have
in
the
bigquery
system
the
very
detailed
revenue
information
that
you
can
slice
and
dice
and
drill,
and
you
receive
the
cost
that
you
can
then
map
together.
So
you
have
a
very
detailed
p
l
without
you
know
the
expensive
part
of
the
s4
hana
system
and
still
with
the
right
level
of
information.
C
So
if
you
have
never
heard
of
the
gcp
query,
I
highly
you
know
recommend
that
you
take
a
look
and
see
what
it
could
do
for
you
by
combining
the
you
know
the
options
with
the
sap
solutions,
all
right.
So
what's
the
outcome
right?
Well,
first,
standardization
right:
that
was
the
requirement
standardized
time
guys
standardized
get
back
to
out
of
the
box
standard.
So
all
of
those
processes
that
we've
been
working
on
assets,
fp,
a
cash-
and
you
know
international
controllership,
ap
procurement.
C
All
of
those
are
not
following
sap
standard
processes,
sap
standard
system.
So
it's
it's
all
good.
The
result
is
that
we
reduce
our
level
of
effort,
especially
we
have
a
very
high
level
of
use
on
business
process
outsourcing
and
a
lot
of
what
they
were
doing
is
really
you
know,
get
connect
to
one
system
extract
data
connect
to
another
system,
extract
another
set
of
data,
reconcile,
find
the
differences
and
then
fix
system
a
or
six
and
b.
C
So,
like
a
lot
of
you
know,
required
work,
but
that
was
not
really
providing
a
lot
of
value
added
for
the
overall
process.
It
was
really
just
patching
our
lack
of
integration
and
data
quality,
so
that
was
very
important
to
us,
then,
because
of
that,
we
were
able
to
reduce
the
number
of
manual
steps
and
the
effort
for
that
piece
of
the
work
as
well
right.
So
less
systems
less
integrations
better
data
quality.
Of
course
your
effort
is
going
to
be
reduced,
so
we
saw
automatically
some
improvement
more
granular.
E
C
Accessible
data,
of
course,
right
we
have
simplified
the
the
the
information
and
now
we're
sharing
the
same
data
platform,
increase
the
end
user
satisfaction,
especially
around
our
finance
community
community,
increase
their
productivity
because
they
don't
waste
time
on.
You
know,
manual
efforts
and
errors
and
then
tighter
system
integration.
C
I
think
this
is
really
the
picture
of
what
you
we
should
be
trying
to
do
in
in
in
our
projects.
I'm
coming
back
here
to
change
management,
because
this
is
something
that
is
important.
We
know
it's
important,
it's
very
hard
to
evaluate
at
least
how
hard
it's
going
to
be,
and
in
particular
for
us,
because
we
switch
from
one
system
to
the
other,
from
oracle,
to
sap,
from
one
language
to
the
other,
from
processes
to
others.
C
It
was.
It
was
harder
than
we
expected.
Let's
be
honest
there,
so
the
project
required
a
lot
more
change
management
effort
than
we
initially
anticipated.
So
we,
you
know,
enhance
the
team
and
usually,
when
you
do
a
cm,
there's
obvious
stakeholders
that
you
want
to
take
into
account.
You
know
the
employees,
all
the
finance
people,
obviously
they're
going
to
have
to
be
involved.
The
project
team,
of
course,
is
involved,
but
we
also
have
indirect
employees
right.
If
you
think
about
it.
C
Google
is
a
hundred
and
twenty
thousand
direct
employees
hundred
and
thirty
thousand
indirect
employees.
I
mean
those
might
not
be
the
last
numbers,
but
the
idea
is
that
we
have
more
indirect
employees,
vendors
and
contractors.
Then
we
do
have
you
know
direct
employees
on
the
payroll
so
very
important
to
get
them
involved.
C
Of
course
we're
changing
the
the
procurement
system,
so
the
vendors
have
to
be
informed
right
because
we're
changing
cost
centers
we're
changing
company
code
numbers,
so
their
invoices
will
have
to
change.
So
we
need
to
get
them
involved
as
early
as
possible.
To
say:
hey,
you
know,
be
careful,
we're
going
to
change
a
few
things
here.
So
so,
let's
now,
let's
get
to
know
it.
E
C
As
for
expertise
like
the
recruiters,
couldn't
make
the
difference
between
those
different
different
attributes,
so
we
even
have
to
get
them
up
to
speed,
so
you
can
think
about
it.
That's
a
lot
of
people
that
have
to
be
touched
and
up
skilled
and
trained,
so
how
we
did
it
well,
there
was
an
internal
website
creating
newsletters.
C
C
Readout
sessions
were
really
critical,
so
the
idea
is,
you
know,
as
we
were
going
through
the
decisions,
we
had
a
number
of
days
that
were
blocked
to
just
read
out
those
decisions
and-
and
then
you
know
a
few
points
in
the
time
where
we
did
next
snacks
and
facts,
so
you
know
bring
people
together
and
then
just
see
what's
going
on,
what's
going
to
change,
what
does
that
mean
for
me?
So
those
were
very,
very
important.
C
C
C
You
design
with
your
you
know,
consulting
partners,
the
perfect
process,
the
ideal
process,
what
we
should
be
doing
and
then
you
implement
it
in
the
system.
You
test
that
it's
all
good,
but
then
people
start
using
the
system
and,
of
course,
they're
going
to
use
it
very
differently,
but
it's
very
hard
to
know
what's
going
on
in
the
system,
so
we
want
to
implement
process
mining
potentially
also
task
mining
and
then
combine
this
with
process
automation.
C
C
So,
instead
of
typically
using
100k,
let's
say
they
put
a
million
dollars
because
it's
you
know
one
zero
different
and
it's
still
within
the
boundaries
of
what's
you
know,
is
allowed
the
transaction
can
go
through
and
then
you
know
it
takes
time
to
figure
it
out
and
by
the
time
you
identify
it.
There
might
be.
You
know,
data,
conversions
or
currency
conversions,
so
it's
hard
to
get
back
to
it.
So
there's
a
number
of
tools
that
you
can
put
together
or
on
the
system
that
look
at
transactions.
C
Those
are,
you
know,
very
process,
heavy
tasks
that
we
do
very
frequently
and
if
we
can
automate
them
and
improve
on
them
through
ai
and
ml,
that
would
be
a
big
benefit:
fraud,
prevention
and
smart
contracts.
I
think
that
we
are
evaluating
as
well
and
see
if
they
could
be
useful
for
us
all.
Right
now,
we
did
benefit
from
a
lot
of
co-innovation
work
with
sap
and
we're
trying
to
actually
you
know,
put
it
back
in
the
community
so
that
everybody
can
benefit
from
it.
C
C
What
we
do
is
we
use
a
feature
in
gcp
call
pops
up
for
publish
and
subscribe
and
all
the
boundary
systems
that
we
have.
They
can
just
connect
to
these
hooks
and
receive
notification
when
there's
a
new
piece
of
master
data
that
they
care
about
so
and
then
they
can
already
do
something
with
it,
maybe
replicate
the
data
or
update
when,
if
it's
an
update,
so
there's
a
lot
of
integration
pieces
that
could
facilitate
your
work
with
other
systems,
integration
with
machine
learning.
Of
course,
now
sap
has
a
lot
of
data.
C
C
So
in
summary
here
and
what
you
could
you
know,
I
think,
learn
from
our
experience.
You
want
to
start
with
core
principles.
How
do
you
make
decisions?
How
do
you
agree
on
anything
that
might
be
simple
to
say,
but
unless
you
have
the
right
priorities
like
do
you
want
to
go
fast,
or
do
you
want
to
do
the
right
thing
sometimes,
depending
on
the
project?
Those
might
be
conflicting.
C
So
in
our
case,
we
really
wanted
to
make
sure
that
we
don't
want
check
that
going
back
to
standard,
for
instance,
was
something
that
was
very
important
to
us,
and
that
was
a
core
principle.
When
we
had
to
make
decisions,
you
can
combine
s4hana
on
ccp,
combine
it
with
s4
public
cloud
and
central
finance,
and
then
you
can
use
them
as
stepping
stones
for
your
finance
transformation
program.
C
What
was
maybe
surprising
for
some
of
our
team
members
was
that
you
know
system
implementation
today
has
less
to
do
with
coding,
then
configuration
especially
if
you
go
back
to
standard.
If
you
think
about
it,
if
you
go
back
to
standard
there's,
not
a
lot
of
you
know
ryceps
and
custom
code,
it's
a
lot
going
to
be
configuration
right.
So
you
it
changes
a
lot
of
things.
For
instance,
you
might
hire
less
developers
and
more
system
analysts.
C
C
So
it's
all
standard
tools,
but
when
you
put
them
together,
it's
very
rare
to
put
them
this
way.
Together,
I
insisted
on
change
management.
I
think
this
is
clear.
Now
it's
very
important
topic
when
you
make
decisions,
it's
a
lot
easier
to
have
a
smaller,
dedicated
group
to
look
at
the
problem.
Come
up
with
options,
recommendations
and
then
go
back
to
the
bigger
group
instead
of
having
everybody
put
their
hand
in
there
in
the
kitchen.
E
C
But
if
you
have
the
right
relationship
with
the
sap
product,
you
can
do
a
lot
and
it's
very
priceless
right.
It
gives
you
really
gets
you
to
a
high
new
level
of
you
know,
results
and
for
us
I
think
our
consulting
partners
were
really
surprised
about
how
a
way
of
asking
tough
questions
and
challenging
the
status
quo
and
building
future
pro
examples
here
are
when
we
were
looking
at
the
customer
design.
C
C
A
A
What
we
have
is
basically
what
you
see
on
your
screen
is
a
subset
of
sap
customers,
all
referenceable
who
are
already
running
on
google
cloud
today.
Right
and
you
know
one
of
the
one
of
the
things
is.
You
know
I've
engaged
with
with
a
few
of
them
over
here
right.
So,
for
example,
you
see
ocean
network
express
which
is
based
in
singapore.
A
You
see
sig,
which
is
seven
indonesia
group
logistics
based
in
indonesia,
bluebird
group
again,
you
know
based
in
indonesia
they're
all.
Basically,
you
know
customers
who
are
running
their
sap
applications
on
on
google
cloud,
and
you
know
every
time
the
question
is
asked.
Why
did
you
actually
choose
google
cloud?
I
think
the
answers
are
very
similar
to
what
julian
has
actually
narrated
right.
A
But
what
we
actually
like
about
google
cloud
is
the
fact
that
you
know
we
are
able
to
leverage,
probably
the
best
cloud
data
platform
that
exists,
and
we
are
able
to
kind
of
you
know,
extend
the
benefits
of
these
business
processes
in
the
form
of
transformation
that
we
require
into
being
a
data
driven
enterprise.
So
almost
all
these
customers
that
you
actually
see
on
the
list
over
here
not
only
run
their
sap
applications
on
google
cloud,
but
they
most
definitely
are
making
investments
into
our
data
infrastructure
as
well
right.
A
So
you
heard
the
term
bigquery.
Of
course
it's
a
little
more
than
bigquery.
This
is
basically
you
know,
technology
that
that's
powering
all
of
the
google
products
right.
So
it's
literally
our
way
of
actually
you
know
bringing
in
technology.
That's
managing!
You
know
all
the
data
that
that
you,
that
kind
of
you
know
end
up
on
on
google
searches
to
to
be
able
to
kind
of
leverage
by
enterprises
for
for
searching
across
their
enterprise
white
data
itself.
Right.
A
So
again,
I
guess
it's
just
a
testament
to
the
fact
that
you
know
the
names
you
see
over
here.
If,
if,
if
google
cloud
is
a
platform
that
you
would
like,
you're
interested
in
in
kind
of
running
your
sap
applications
with,
I
really
genuinely
hope
and
pray
that
you
also
consider
you
know
being
data
driven
enterprise
as
being
part
of
your
entire
chapter
itself.
Right,
if
you
can
move
to
the
to
the
next
slide,
julia.
A
I'm
not
going
to
dive
into
the
details
of
this.
I
just
realized
that,
yes,
there
are
some
sap
technical
folks
on
this
call
as
well.
This
is
a
reference
architecture.
It's
a
pretty
detailed
architecture.
This
is
again
published
on
the
on
the.
If
you
go
to
the
google
cloud
public
website
itself,
you
know,
I'm
happy
to
kind
of
you
know,
share
that
link
on
the
chat,
also
where
you
can
actually
find
all
the
public
documentation
right.
A
But
what
I
want
to
kind
of
assure
is
that
we
we
we
have
sap
certified
machines.
We
have
a
fully
certified
reference
architecture
that
we
work
with
sap,
which
supports
you
know
high
availability,
which
supports
you,
know,
disaster
recovery,
and
you
know
we
are
happy
to
actually
have
a
much
more
detailed
conversation
on.
You
know
how
this
might
actually
look
for
for
your
environment
itself
right.
A
So
that's
just
you
know,
I'm
not
getting
good
getting
into
details
of
this
just
want
to
assure
you
that
we,
we
have
a
very,
very
robust,
solid
architecture,
leveraging
some
of
the
best
services
within
google
cloud
to
implement
a
highly
available
and
a
disaster
resilient
sap
solution
on
google
cloud.
All
right
with
that,
I'll,
probably
you
know,
have
rohan
actually
go
through
the
rest
of
the
content.
A
B
Yeah
thanks
narayan,
thanks
julian,
so
hello,
folks,
so
so
till
now
we
have
heard
about
the
kind
of
transformation.
What
we
have
done,
google
wide
for
our
financial
processes,
as
well
as
other
processes
so
and
naren,
also
touched
about
our
certification
that
we
have
done
with
sap
to
give
infrastructure
for
sap
or
even
other
services.
But
if
you
are
evaluating
google
cloud
from
india
perspective,
this
is
this
is
the
slide
for
you?
Okay,
so
first
is
currently.
B
We
are
the
only
hyperscaler
in
india
who
has
two
regions
in
two
different
seismic
zones,
and
each
region
has
multiple
availability
zone.
Now
availability
zone
is
like
a
failure
domain
failure
domain
in
the
sense,
if
one
failure
domain
goes
down,
that
can
be
power
that
can
be
cooling
or
any
other
infrastructure
issues.
The
other
failure
domain
is
insulated
from
that
failure
right.
So
that's
what
that's
what
the
availability
zone
is
all
about.
B
So
currently,
google
is
the
only
hyperscaler
in
india
with
which
you
can
have
all
sorts
of
these
architectures,
which
are
mainly
critical
for
sap
erp
systems
in
india
right
now,
the
other
differentiator
again
from
india
perspective.
What
we
have
is
currently
google
cloud
is
the
only
hyper
scaler
where
customer
a
live
customer
is
running
12
tb,
virtual
machine,
so
20b
is
for
for
hana
and
it's
again
comes
in
a
virtual
form.
The
benefit
of
having
consuming
this
resource
in
a
virtual
is,
you
can
manage
it
with
code.
B
There
are
some
marquee
customers
who
have
selected
google
cloud
for
the
recipe
deployments,
and
I
just
picked
a
subset
of
this
just
to
highlight
that
there
are
one
of
the
largest
conglomerates
in
india,
not
just
in
india,
even
in
bangladesh.
They
are
running
their
erp
on
google
cloud
right
now.
What
is
it
that
we
offer
as
a
google
cloud?
If
I
want
to
summarize
in
one
single
slide,
what
are
the
differentiators
or
what
are
the
value
that
we
add,
apart
from
the
information
what
you
heard
from
julian?
So
this
is
one
single
slide.
B
So
first
point
is
secure
network
right,
so
google
has
built
a
dedicated
network
for
google
itself.
It's
a
very
deep
and
high
throughput
fiber
network.
What
we
are
laid
out
throughout
the
globe-
and
it
is
it
you
can
it's
a
very
time
tested
and
battle
hardened
network,
because
it's
the
same
network
that
is
used
by
youtube.
Google
search,
all
the
google
public
facing
utilities
or
tools
that
we
use
it's
based
on
the
same
network
right
and
there
are
other
subtle
nuances
of
these
networks
as
well.
B
So
typically,
whenever
for
sap
systems
generally,
you
might
have
seen
shapes
like
8
vcpu,
64,
gb
ram
or
4
vcpu
32gb
ram
like
it
comes
in
typical
shapes,
but
if
your
workload
demands
a
very
odd
type
of
shape,
let's
say
6
vcpu
and
73
gb
ram.
Now
this
is
not
a
standard
configuration
which
is
available
out
of
the
w,
but
in
case
of
google
cloud
you
can
do
that
customization
so
that
you
are
only
deploying
infrastructure
which
is
required
for
your
workload,
and
you
are
only
paying
for
that.
B
So
that's
a
one
step
above
how
we
can
go
pay
as
you
go
concept,
the
third
and
most
important
again,
it's
a
unique
differentiator
to
google
cloud
is
we
call
it
as
a
live
migration,
so
live
migration
is
a
concept
where
we
have
algorithms
in
our
data
centers
they
keep
on
tracking
the
infrastructure
in
backend.
So
if
they
realize
a
particular
infrastructure
is
going
to
fail
in
some
time
proactively,
it
will
lift
your
vm
virtual
machine,
which
is
running
on
top
of
that
infrastructure,
and
it
will
shift
it
to
a
healthy
host.
B
Now
this
lifting
and
shifting
that
happens
in
a
live
way.
That's
why
we
call
it
as
a
live
migration.
So
during
this
transaction,
as
in
during
this
live
migration
event,
your
end
users
can
still
continue
to
work.
They
won't
feel
any
disruption
or
neither
you
will
get
a
notice
from
google
saying
that
we
are
doing
live
migration.
You
need
to
take
a
down
time
so
because
of
which
we
can
maintain
our
hardware
without
you
taking
any
downtime,
so
that
improves
the
native
high
availability
out
of
the
box,
and
there
is
no
extra
charge.
B
No
extra
configuration
nothing.
It
is,
it
is
baked
in
into
the
infrastructure
by
default,
for
you
and
the
last
and
foremost
which
again,
I
would
like
to
build
on
what
julian
has
just
said
about
innovation,
so
google,
internally
google,
has
a
dna
of
innovation
and
we
try
to
bring
in
more
and
more
products
which
can
solve
industry
use
cases,
but
with
little
bit
innovation
differently.
Now,
as
you
know,
julian
has
talked
about
multiple
products
like
big
query:
pops
up
how
we
can
connect
this
all
this
together.
B
But
apart
from
that,
we
have
also
gone
one
step
ahead
in
ai
and
as
well
as
machine
learning
where,
even
if
you
don't
know
how
to
write
machine
learning
code,
if
you
don't
know
r
language
python,
you
can
still
achieve
machine
learning
by
writing.
Sql
queries
right
and
again
that
works
well
with
sap
as
well.
So
these
are
the
four
points
which
I
think
that
you
know
google
cloud
can
help
you
with.
If
you,
if,
if
you
are
evaluating
google
cloud
for
your
cloud
journey,.
B
D
Thanks,
thank
you
thanks,
one
for
for
doing
that.
I've
started
a
poll
just
so
that
we
understand
whether
the
session
today
was
useful
or
and
if
you
would
like
to
know
more
about
topics
like
this
in
the
future
as
well.
So
while
the
polls
kick
in,
I
think
let's
take
one
or
two
questions.
I
know
that
we
are
already
on
top
of
the
hour.
D
So
if
someone
wants
to
drop
off,
please
feel
free
to
do
so
for
people
who
have
asked
the
questions
and
I'm
assuming
that,
can
the
google
team
stay
online
for
a
few
more
minutes,
sure
yeah,
absolutely
okay.
So
the
first
question
comes
is
around
mna,
so
maybe
amit.
If
you
can
also
answer
this
question
from
an
acv
point
of
view,
so
there
is
a
customer
who
says
that
you
know
there
is
going
to
be
a
lot
of
m
a
activity
that
they
are
going
to
experience
in
the
near
term.
D
C
C
Very
different
type
of
business,
if
you
do
vertical
integration
right
now,
in
any
case,
you
have
a
legal
requirement,
at
least
in
the
u.s
and
in
europe
in
within
90
days.
You
need
financial
consolidation
of
the
the
new
entity
to
get
to
90
days.
You
don't
have
90
days
to
implement
sap
right.
So
what
you
typically
do
is
a
very
high
level
integration,
consolidation
and
therefore.
C
Is
to
go
directly
to
bpc
or
group
reporting
or
whatever
your
consolidation
system
is
right
so
right
now
the
the
quick
and
dirty
solution
is
to
only
do
financial
consolidation.
What
you
want
to
do
is
align
at
the
company
code
level,
potentially
division
and
whatever
you
report
on
right,
so
strategic
first
step
finance
consolidation.
C
Now
from
then
on,
you
could
stay
there
like,
because
when
you
do
mna
you
might
be
buying
and
selling
companies
right.
So
that's
the
first
you
now!
If
you
do
this,
you
are
now
compliant
with
the
law.
Now
you
don't
get
all
the
benefits
we
want
right.
So
in
order
to
get
the
benefits,
you
need
to
integrate
the
processes
and
to
integrate
the
processes.
A
C
You
want
to
have
you
know
one
if
the
if
the
customers
are
similar
right,
you
buy
your
choir
company
within
the
same
country
with
a
slightly
different
business
use
case,
so
if
they
have
the
same
customers,
you
want
to
have
a
single
point
of
credit
check.
To
do
that,
you
need
integration
at
the
finance
level.
C
And
you
can
do
just
all
the
finance
pieces
and
central
finance
helps
a
lot
with
that.
It
doesn't
matter
if
the
if
the
acquired
company
is
sap
on
an
sap
all
right,
then,
if
you
really
want
to
have
a
full
integration,
then
you
have
two
options:
either
you
merge
the
acquired
company
into
the
acquiring.
E
E
C
Get-Together
or
you
could
do
some
processes
with
central
finance,
not
all
of
them,
but
some
of
them
or
then
you
have
to
implement
the
whole
erp
right.
So,
for
instance,
let's
say
google,
which
company
has
you
know
devices
services
we
have
nests.
We
have
smart
watches
that
we
acquired
from
fossil
a
while
ago,
and
we
just
acquired
fitbit
if
we
want
to
have
a
merged
strategy
to
do
a
single
warehouse,
for
instance,
or
a
single
sales
team
that
goes
out
to
best,
buy
and
sells
both
of
those
equipments
together.
C
You
need
to
be
in
the
same
erp
system
or
you
need
to
integrate
with
central
finance,
but
you
might
have
limitations
in
what
you
can
do.
For
instance,
you
would
have
obviously
different
plans
that
you
have
to
articulate
between
the
two
systems
right,
so
it
it
gets
a
lot
harder
if
to
be
in
different
erp
systems.
If
you
don't
have,
if
you
want
to
have
a
full
process
integration-
and
I
realized
it
was
a
one
second
question-
it
was-
was
a
five
minute
answer,
but
I
hope
this
is
helpful.
E
No
book
is
nothing,
nothing
to
it,
but
again
you
know
from
amanda's
perspective.
Yeah
one
part
is
to
get
the
numbers
at
one
place,
and
then
you
know
second
part
is
how
you
can
operationalize
the
data
as
quickly
as
possible.
So
this
is
where
you
know
central
finance
comes
into
picture
and
I
think
a
google
perspective.
They
are
using
more
on
the
vpc
site
right,
but
we
do
have
a
group
reporting
solution
which
merges
both
layers
into
one
picture.
E
A
D
C
That's
a
great
question
so
so
we
had
to
define
one
of
those
decisions
was
actually
historical
data
and
what
did
we
do
in
there
right?
For
us?
It
was
not
an
sap
system
upgrade
right.
We
had
a
green
field
implementation,
so
we
started
from
an
empty
box.
So
when
you
start
from
an
empty
box,
you
have
to
decide
how
much
do
I
carry
over?
Do
I
carry
over
one
year,
three
years,
five
years
or
everything
right
so.
E
A
C
A
E
C
C
C
C
The
rest:
well,
it's
it's
hard
for
us
as
google
to
say
we're
going
to
get
rid
of
it
right.
It's
like
you
know.
Google
loses
data,
it's
kind
of
a
problem
right
so
and
we're
constantly
under
scrutiny.
So
we
we
always
have
questions
going
back
to
10
or
15.
Now
there
was
even
recently
a
request
back
to
20
years
ago
right,
like
an
asset
you
might
have
acquired
or
an
employee.
You
might
have
paid
or
something
like
that.
So
so
we
have
to
keep
our
old
system
on
ice.
C
Basically,
you
know
turn
it
down
as
low
as
possible
and
and
keep
it
that
way.
So,
basically,
three
levels
of
strategy
active
data
in
the
s4
system
a
year
plus
right
a
year,
and
then
you
start
bw
system
is,
I
think,
seven
years
and
then
the
rest
stays
in
the
old
system.
The
way
it
is
in
terms
of
volume.
I
cannot
share
that
data,
but
the
records
we
have.
If
you
can
think
about
it,
we
any
eight
million
finance
records
per
month
and
that's
just
finance.
E
E
C
The
very
technical
questions
here,
but
the
way
we
implemented
it
is
that
we
replicate
the
bw
data
into
bigquery.
The
reason
for
this
is
that,
if
imagine
you
had
a
direct
connector
from
bigquery
to
hana,
you
would
have
more
load
on
your
bw
system.
That's
not
what
we
want
right.
We
want
to
push
the
load
to
another
system
right,
but
maybe
narian
and
ohani.
If
you
have,
if
you
have
more
details
here,.
A
Sure
sure,
no,
so
so
the
the
way
we
actually
look
at
it,
but
I
mean
I
I'll
probably
just
give
the
example
of
some
of
the
customers
that
I
spoke
about
like
home
decor,
for
example,
right
home
depot
runs
bwbbc,
just
like
we
do,
but
their
use
of
bigquery
actually
goes
beyond
just
say.
The
finance
domain
itself
right
for
them.
Bigquery,
essentially
is
kind
of
you
want
to
call
it
a
data
warehouse.
A
And
also
a
lot
of
iot
data
as
well
right.
So
the
way
we
look
at
it,
you
know
what
the
analytics
that
that
basically
require
recipe
data
and
all
of
that
is
kind
of
available
within
your
within
your
sap
systems
itself.
The
way
customers
actually
prefer
to
implement
bigquery
is
to
complement
what
you
already
have
from
an
sap
solution
offering
perspective
and
to
actually
bring
in
a
lot
of
you
know:
iot
data
non,
probably
data,
that's
coming
out
from
you
know,
non-sap
applications
and
bigquery
essentially
becomes
a
repository
where
you
can.