►
From YouTube: Digital Transformation of Vision Banco Paraguay with Serverless... - Alex Ellis & Patricio Diaz
Description
Join us for Kubernetes Forums Seoul, Sydney, Bengaluru and Delhi - learn more at kubecon.io
Don't miss KubeCon + CloudNativeCon 2020 events in Amsterdam March 30 - April 2, Shanghai July 28-30 and Boston November 17-20! Learn more at kubecon.io. The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy, and all of the other CNCF-hosted projects
Digital Transformation of Vision Banco Paraguay with Serverless
Functions - Alex Ellis, VMware & Patricio Diaz, Vision Banco SAECA
This talk explores how a bank in Paraguay struggled with fire-fighting issues with their monolithic enterprise service bus code and embarked on a journey of digital transformation with serverless functions and Cloud Native technology. For Vision Banco SAECA Serverless with OpenFaaS gave an easy on-ramp to start making use of technology like Docker, Kubernetes and Prometheus within days through user-friendly abstractions and workflows. Patricio Diaz who leads the effort at the bank is now running two key parts of the home banking service (HomeServe) with OpenFaaS functions. Due to the sensitive nature of the workloads - monitoring dashboards are important for the team and these were set up with the existing Prometheus data and Grafana. Given the initial success Patricio is working on moving more of the existing system functionality over to functions.
To learn more: https://sched.co/GraO
A
Hello,
everybody
knocking
on
the
door
and
let's
get
started
with
the
session,
so
I'm
here
with
open,
FAR's,
end-user
and
contributor,
patricia
he's
flown
in
all
the
way
from
Paraguay.
Just
to
speak
to
you
today,
and
my
name
is
Alex
Ellis
and
I'm
the
creator
of
open
farce
and
work
it
VMware.
So
we
want
to
take
you
on
a
bit
of
a
journey
today
and
what
I've
done
here
is
I've
put
up
a
globe
and
right
up
in
that
corner
is
Washington
State,
which
is
where
we
are
right
now.
A
Thousands
of
miles
eight
hours
nine
hours
across
here
is
United
Kingdom,
where
I
live
and
Patricia
from
South
America.
Here.
Why
have
I
put
this
up
on
the
screen?
Well
with
thousands
of
miles
apart,
but
we
have
exactly
the
same
technical
problems
and
this
community
of
kubernetes
and
cloud
native
components
is
able
to
help
us
solve
these
problems,
and
we
actually
find
that
our
community
is
got
closer
by
resolving
these
issues
and
focusing
on
end
users.
A
The
first
contact
I
had
with
Patricia
was
actually
this
tweet,
where
he
said:
I've
just
launched
an
open,
fast
function
into
production
and,
as
per
normal
I
thought.
Oh,
that's
great.
Until
I
found
out
that
he
was
actually
a
bank
in
Paraguay
and
he
was
running
service
functions
in
production
using
open
pass,
and
so
that's
where
a
relationship
started.
She
wouldn't
say
a
few
words
well
yeah.
B
Before
I
hi
to
everyone,
I'm
Patricia
yeah,
at
that
time
we
had
our
function
working
with
his.
They
were
three
functions
or
prototype,
we're
working
for
about
a
week
with
all
a
problem
and
it
work
it
for
us.
It
was
a
better
better
scenario
than
we
have
with
with
that
same
processor
with
or
our
previous
infrastructure.
Now.
A
B
A
A
How
you
guys
can
make
use
of
this
and
take
the
knowledge
back
to
your
workplace,
how
it
compares
to
other
service
frameworks
are
available
and
then
we'll
look
at
some
of
the
lessons
that
we
learned
and
what
we're
going
to
do
in
the
future.
As
we
look
forward,
you
see
the
bottom
of
the
slide.
Our
twitter
handles
feel
free
to
take
photos
and
post
them.
Our
contact
details
will
be
there
at
the
end
as
well
have
a
chance
to
ask
questions.
Okay,
so
tell
us
a
bit
about
Paraguay.
Well,.
B
I
am
from
far
away.
This
is
a
country
north
of
America.
We
are
an
island
surrounded
by
chrome.
We
like
to
say
I,
want
to
to
tell
you
true
to
remarkable
facts.
We
have
just
a
reference.
The
first
one
is:
we
have
free
hydroelectric
dams.
The
hundred
a
hundred
percent
of
the
electricity
we
use
on
our
country
comes
from
those
dams.
One
of
them
is
the
second-biggest
done
in
the
world.
Well,
the
other
interesting
fact
is
that
we
exported
and
soya
beans,
if
probably
the
soil
aid
you
you
like
to
drink.
B
B
That
mean
that
means
the
euros
us
or
Bank,
as
I
told,
is
ranked
one
or
bangies
ranked
it
at
the
top.
Ten
of
among
the
top
ten
of
the
banks
and
the
country,
as
I
told
one-third
of
the
population
have
accounts,
had
bank
accounts
and
400,000
of
those
customers
are
cousin
of
our
bank
or
bangies
is
focuses
on
small
and
medium-sized
business.
That
is
why
your
goals
inspires.
B
Us
is
most
enterprise
statement
well,
but
that's
changing
a
little
because
our
government
is
doing
campaign
to
make
most
of
the
individuals
or
the
population
become
customers
of
the
financial
system.
Well,
that,
among
that
increasing
amount
of
customers
makes
that
millions.
We
are
system
receive
millions
of
requests
per
day.
Well,.
A
That's
great
tonight,
and
so
I
hope
this
gives
you
a
little
bit
of
a
flavor
of
Paraguay
and
vision
bank
and
now
we're
going
to
start
to
look
at
the
journey.
Now
this
journey
started
off
in
the
enterprise
they've
heard
about
this
enterprise
mission
using
JBoss
from
Red,
Hat
and
JBoss
is
a
framework,
that's
related
to
ESB
fuse
that
allows
you
to
build
micro
services.
A
A
There
Ron
Patricia,
as
we
saw,
adopted,
docker
with
swarm
and
open
fast,
that
was
on
the
tweet
a
little
bit
later
on,
though
in
the
journey
you've
moved
to
Cuba
Nettie's
and
that's.
Why
we're
here
at
cube
con
to
tell
you
how
kubernetes
opened
fast
and
related
technologies?
It
can
build
a
complete
solution
that
solves
problems.
Okay,
so
tell
us
what
it
was
like
with
JBoss
well.
B
We
started
using
gables
and
before
other
application
servers,
but
lately
we
are
using
ebooks.
We
we
use
in
him
because
he
has
a
very
good
enterprise
support.
We
benefit
from
the
composition,
but
but
we've
lady
with
foam,
also
that
they
are
some
drawbacks.
It
is
us
which
is
a
little
Bridal
as
well.
It
is
unstable
on
the
skull
is
very,
very
resource
hungry,
so
it
is
very
hard
to
skull
to
skull.
We
had
to
babysit
it
a
lot
and
we
had
to
have
people
all
the
time
watching
it
and
taking
care
of
it.
B
Seeing
the
scenes
from
the
developers
point
of
view,
it
is
an
complex
architecture
to
carry
on
single
project.
We
had
to
gather
a
lot
of
people
a
lot
of
stakeholders
and
hope
and
wait
that
they
can
agree
the
concerns,
responsibilities
and
requirements
of
the
project
and
that
that
takes
prolong
a
very
long
time
and
it
is
prone
to
to
fail
that
doesn't.
A
Sound
very
cloud
native
to
me:
I,
don't
know
about
you,
so
let's
look
at
what
we
were
able
today
to
help
out.
So
the
first
step
in
this
journey
was
to
start
looking
at
functions
and
with
that
you
actually
got
a
lot
of
the
docker
ecosystem,
so
self-healing
declarative
infrastructure.
Now,
when
we
look
at
this
architecture
of
open
fires,
we
have
an
API
gateway
component.
We've
got
a
watchdog
that
is
similar
to
what
Kelsey
showed
you
in
the
keynote
this
morning
can
make
any
function
service
any
binary
and
then
Prometheus
metrics.
A
That
allows
the
monitor
the
system
and
that's
streaming
to
do
asynchronous,
processing
and
a
very
lightweight
level
at
a
very
low
cost,
and
then
we
launched
with
swarm.
But
now
we
have
native
Cuban
IT
support
that
just
runs
on
docker
images
and
it's
a
key
differentiator
between
other
platforms.
What's
the
most
important
part
of
this
to
you
and
what.
B
A
B
A
That's
coming
from
see
advisor,
but
actually
you
need
a
bit
more
than
just
Bram
and
CPU
and
that's
why
I
open
fast
as
custom
metrics,
so
it
collected
at
that
gateway
component
that
allow
us
to
do
this
auto
scaling
to
whatever
needs
you
have
and
even
to
zero
and
back
again,
and
so
we
have
a
demo
here
where
this
was
taken
during
docker
con.
We
had
a
huge
calling
on
this
function.
A
There
was
replica
scaling
up
and
up
and
up
and
then,
when
that
request
load
had
passed
its
go
back
down
to
zero
again,
so
you
were
able
to
track
these
metrics
and
for
remember
back
to
that
tweet
Patricia
cut
out.
It
was
a
graph
on
a
dashboard,
so
it's
very
important
to
be
able
to
monitors
Telus.
Just
very
briefly
about
you,
service
journey
with
swarm
and
open
vests
yeah.
B
Well,
we
are
not
that
open
path
and
once
we
have
defined
the
requirements
and
and
the
requirement
for
a
project-
and
we
can
start
doing
functions
and
it
takes
us
about
a
few
hours
and
two
or
three
days
a
week
at
at
maximum
to
have
I'm
just
working
to
have
a
function
working
doing
the
things
with
natural
natural
natural
functions
are
very
resource
efficient,
so
we
are
not
waiting
wasting
resources
and,
as
the
platform
gives
us
control,
also
healing,
auto
scaling
up
and
down.
We
we
had
to
do
a
less
babysitting
with
or
deployment.
A
B
A
B
What
we
found
we
visited,
it
was
easy
to
install.
We
just
started
it
without
a
problem.
Then
we
started
open
pass
over
it
and
we
start
using
it
as
we
were
used
to
to
do
it.
So
it
was.
It
was
not
a
big
deal
to
move
from
2
to
kubernetes
and
open
fast
opener.
We
have
not
more
control
of
visibility
and
visibility.
B
With
with
that
now,
when
we
had
successful,
running
or
function
with
open,
paws
and
kubernetes,
we
need
to
find
a
tool
to
help
more
control,
to
help
visibility
and
to
allow
our
deters
and
tester
and
and
that
kind
of
people
to
to
manage
or
cluster.
So
after
a
couple
of
candidates,
we
found
rancher,
we
found
it
easy
to
itself.
We
found
it
very,
very
easy
to
use.
You
have
pretty
good
UI,
but
most
most
important.
B
B
Now
we
had
to
do
with
more
lines
or
more
steps,
but
just
kind
to
get
used
to
it
and
other
drawback
we
found
we
found
a
within
help
with
swarm
is
that
we
now
had
to
use
service
discovery,
and
it
is
easy
to
use
with
all
functions
that
all
the
project
that
I
deploy
and
coordinates,
but
we
had
to
integrate
all
that
with
our
all
infrastructure.
So
we
we
are
now
embarking
to
to
implement
and
learn
how
to
use
service
discovery.
A
B
We
are
trying
to
lead
the
the
technology
innovation
in
Latin
America.
We
are
trying
to
do
that.
We
are
contributing
to
open
source,
maybe
not
with
much
code,
but
we
but
doing
use
case
and
giving
heat
but
feedback
to
the
community
and
that's
very
important
to
taking
in
a
counter.
We
are
a
very,
very
restrictive
environment,
to
run,
to
run
we're
a
bank
yeah
and
we
adopted
core.
Ladies,
we
are
using
it.
B
A
Yeah,
nobody
forget
that
so
a
lot
of
the
work
that
Patricia
has
done
and
a
lot
of
the
value
that
he's
drawn
from
is
from
this
landscape.
Now
this
is
the
serverless
landscape.
You
all
very
familiar
with
the
cloud
native
one
and
on
the
left-hand
side
of
this
screen.
Believe
it's
the
Left
we've
got
hosted
platforms
that
do
suit
a
lot
of
people's
needs,
they're
completely
locked
down.
You
have
very
little
control,
but
they're
very
easy
to
use
on
the
right
hand
side
in
this
box.
You
have
installable
service
platforms.
A
Now
these
are
a
completely
open
source
and
two
logics,
then
you're
in
control.
You
can
get
under
the
bonnet
and
you
can
make
adjustments
as
you
need
to
numb
into
your
team
now
for
Patricia.
Oh,
he
needed
something
that
he
could
run
on
premise
you
may
have
multi
filed,
you
may
have
hybrid,
but
they
have
private
club
only
so
they
needed
something
from
this
landscape
and
they
picked
open
fares,
but
I'd
like
to
highlight
dispatch
from
VMware
and
never
internal
effort
project.
A
Rif
KNAT
we've
heard
a
little
bit
about
that
at
the
conference
and
then
there's
some
other
big
projects
there
too.
Now
these
will
have
different
pros
and
cons
and
I
thought
we
just
walk
through
a
few
of
them.
Give
you
a
bit
of
context
so
open
fires
and
keyless,
probably
a
couple
of
the
most
mature.
Here
we
go
back
in
two
years.
A
They
come
from
the
community,
open
fires,
q+
produced
bitten
army
and
google
recently
reduced,
released,
install
controls,
k
native
okay
native,
is
more
of
a
platform
builders
tool
than
a
service
project
that
you
can
consume
as
a
user
open
path
and
keep
us
have
very
nice
user
interfaces.
You
spend
a
lot
of
time
on
those
they're,
both
very
stable,
ok,
natives,
probably
around
0.2.
If
I'm
correct
and
he's
a
very
very
new
was
released
three
months
ago
by
google
in
the
summer,
whereas
we
have
many
more
customers
in
production.
A
Open,
fast
and
k
native
can
both
run
a
docker
image,
which
means
any
binary
that
you
have
can
actually
run
as
a
function.
Key
plus
limits
you
to
just
the
rent
times
that
they've
picked
and
then,
if
you
look
at
this
column,
it
says
yeah
mall,
so
open
fires
has
a
stack,
yeah
more
file,
which
is
a
really
minified,
simplified
surface
area
that
designed
for
all
of
us,
but
we
also
have
a
CID.
A
Think
when
you
embrace
something
like
that,
you're
taking
on
a
lot
of
ambien
cost,
but
for
certain
reasons
dipping
in
a
bit
more
to
that,
we
have
basically
infinite
code
limits
for
both
of
these
projects
and
things
like
cublas
that
mirror
something
like
lambda
have
a
one
megabyte
limit
for
your
function
because
they
use
a
config
mapping,
Kuban
a
tease,
okay
and
then
these
two
projects
will
scale
to
zero
and
then
back
again
was
a
lot
of
different
projects
in
the
landscape.
That
hasn't
been
a
priority
for
them.
Why
am
I
showing
you
this?
A
A
Okay,
so
open
fire
started
in
November
2016
as
a
docker
captain,
as
a
container
geek
I
was
trying
to
find
out,
could
I
take
that
experience,
I
had
with
lambda
and
improve
it
and
actually
run
it
on
my
own
containers
and
that's
something
that
I
did
and
in
this
picture
you
see
me
at
docker
con
just
around
a
year
ago,
presenting
it
to
5,000
people
and
from
there
it's
just
gone
from
strength
to
strength.
I
have
a
full-time
team
at
VMware
now
where
the
project
remains
independent,
but
I
have
the
resources.
A
I
need
to
develop
this
with
the
community
and
work
with
customers
like
Patricia.
We've
got
people
that
have
got
this
in
production
over
the
last
two
years.
We've
built
up
that
knowledge
with
looking
at.
How
do
we
solve
these
problems,
for
you
guys
and
then
bosie
awards,
so
infoworld
just
awarded
us
a
second
bosie
award
for
best
cloud
computing
software,
we're
really
proud
of
we're
driven
by
values,
rather
than
solving
a
technical
problem
at
that
level
we're
trying
to
be
driven
by
some
driving
values,
one
of
those
is
being
community-centric.
A
A
When
you
have
this
many
people,
you
need
some
kind
of
structure,
some
way
of
getting
them
engaged,
and
so
we
held
regular
Xoom
calls
where
people
can
come
and
voice.
Our
opinions
ask
questions
and
get
involved
and
see
what
we're
building
and
be
part
of
something
bigger,
and
we
also
have
core
contributors
across
the
top
line
and
the
full-time
staff
at
VMware,
so
community
and
organic
independent
community,
something
that's
really
important.
For
us.
Another
value
is
being
easy
to
use
or
putting
developers.
First
now
the
name
of
the
project.
A
The
tagline
is
service
functions
made
simple,
because
that
is
our
focus
and
when
you
come
from
the
cloud
where
you're
typing
text
into
an
editor
and
hitting
save,
we've
got
quite
a
high
bar
to
get
to
right,
and
so
that's
something
that
we
knew
we
had
to
focus
on
from
early
on.
You
can
actually
run
any
code
I
liked
that
Kelsey
used
that
Fortran
example,
because
I
showed
about
a
year
ago,
running
COBOL.
In
the
same
way
a
particle
physicist,
Lucas
her
who
works
at
CERN
came
up
to
me.
A
He
said
we
love
open
fires,
because
we
can
run
our
Fortran
code
in
your
containers
again,
something
that
I
think
we're
seeing
more
and
more
validated,
so
any
kind
of
binary
any
programming
language.
You
can
take
that
to
any
cloud
because
of
kubernetes
and
the
other
support
we
have
and
then
just
because
it's
open-source
doesn't
mean
that
it's
set
to
a
certain
limit.
Kubernetes
again
allows
us
to
to
burst
out
and
have
infinite
scale.
A
So
the
key
differentiator
here
between
a
cloud
function
is
that
you
get
to
decide
what
the
limits
are,
how
long
it
can
run
at
maximum.
How
much
code
you
want
and
all
the
other
parameters,
so
these
are
the
programming
languages
that
we
bundle
as
templates.
Are
there
any
of
these
that
you
make
use
of
at
the
bank
yeah.
B
A
A
This
is
all
you
have
to
look
at.
Is
your
handler
and
some
way
of
saying
what
modules
you
want
for
a
function.
This
is
Python
and
it
goes
off
to
a
website,
pulls
the
resource
and
then
tells
you
what
the
status
code
is.
Is
it
not
found
200?
Okay,
that's
all
that
you
see
as
a
developer
and
then
we
hide
from
you
a
very
well-crafted
well
maintained,
bakka
file
and
entry
point.
That's
got
best
practices
of
light
of
it
so
running
in
a
non
rate
user
in
a
very
tiny
linux
image.
A
Maybe
scratch
if
we're
able
to
health
checks
built
in
for
you.
So
as
I
will
build
your
code
by
combining
that
home
to
learn
the
Python
template
together.
You'll
then
get
a
docker
image
and
it
gets
pushed
to
a
registry.
So
when
you
do
a
fast
Eli,
deploy
or
not
you're,
just
really
deploying
from
a
registry,
and
we
all
know
how
to
build
and
manage
containers,
so
this
is
something
that
works
very
well
with
you.
Instead
of
against.
A
You
know,
when
you've
got
a
couple
of
functions,
you
actually
need
some
way
of
managing
that,
and
so
we've
got
this
lightweight
yeah
more
file.
The
top
one
is
just
three
four
lines
to
define
a
function,
but
the
bottom
one
also
has
a
secret
in
it.
Now
that
same
amount
of
text
in
kubernetes
is
300
lines.
What
would
you
rather
maintain
100
off
right,
I,
think
I
know
the
answer
now:
Patricia,
there's
a
HTTP
proxy
here:
do
you
have
to
use
a
proxy
for
anything?
We.
B
A
A
And
I
mean
it
go
fast,
you'll
I
knew
will
put
language
go
the
name
of
the
function
and
because
this
is
going
to
go
in
the
docker
hub
I'm,
just
putting
the
prefix
of
the
function
that
might
might
username.
This
has
generated
the
code
and
the
Yama
file
that
we
talked
about
so
now.
What
we'll
see
is
our.
A
Okay,
we
have
the
Yama
file,
it
says
what
the
language
is.
It
says
where
it's
going
to
be
deployed,
but
really
what
we
want
to
see
is
the
handler,
and
it
doesn't
really
get
any
simpler
than
that
from
here.
We
can
now
do
fast,
CLI
up
and
in
a
number
of
moments
will
see
that
the
image
was
built
by
docker
locally
has
been
pushed
up
to
the
Internet
and
will
be
deployed
in
the
cluster
I'm,
not
suggesting
you
do
that
in
production,
but
for
development.
This
is
a
very
easy
workflow
in
production.
A
Your
CI
CD
server
would
do
this
and
we've
got
some
other
options
for
you
too.
Now
what
you
see
this
is
in
the
UI
of
open
flat,
a
left
hand,
side,
we've
got
the
functions
in
the
cluster.
There's
our
function
and
it's
been
pulled.
We
can
invoke
it
now,
one
of
the
other
parts
of
open
files.
Ui
is
this
function
store
and
from
here
we
can
get
any
number
of
predefined
functions
out
of
the
community,
store,
deploy
them
and
start
getting
shared
value
from
that.
A
And
I'm
going
to
do
fossilised,
store
list
using
the
command
line.
You
can
see
all
of
these
different
ones
from
the
community
version,
but
actually
you
can
add
your
own
store,
and
this
is
what
we've
done
so
now.
What
I
want
to
do
is
I
want
to
deploy
something
that
looks
a
lot
like
what
Patricia
would
build
at
work
when
your
loan
calculator
and
in
a
web
page
of
form
that
we
can
go
to
to
see
to
see
it
and
access
it.
A
B
A
I'm
gonna
give
you
15%,
cuz
I,
don't
know
about
this
one
and
so
life
that
just
ran
a
function
to
serve
it
and
a
function
to
do
the
projection,
and
as
we
change
the
months,
we
can
see
that
live
being
translated
and
bringing
the
data
back.
So
what
we
saw
was
we
created
a
new
function
very,
very
quickly:
we've
got
it
up
and
running
on
a
cluster
and
Kuban
eighties.
The
custom
function
store
using
stuff
that
makes
sense
for
our
business
now.
Open
VA's
has
a
very
kubernetes
native
experience.
A
We
looked
at
this
earlier,
but
I
want
you
to
focus
on
is
on
the
right
hand,
side
that
open
pass
operator.
This
works
through
custom
resource
definition
called
a
function.
Now
you
define
in
similar
way
that
a
yam
well
that
you
saw
and
then
that
function
gets
deployed
as
an
entry
and
our
operator
will
create
the
pods
that
you
need
it.
The
top
one
has
two,
which
means
it
probably
needed
to
scale
up
because
of
traffic.
A
Okay,
now
make
these
slides
available
for
you,
and
you
can
get
these
online
if
you
want
to
dig
in
a
bit
more
another
thing
that
not
a
lot
of
people
realize
about
open
bars
is
that
it
supports
a
multitude
of
different
events.
For
instance,
one
of
the
most
popular
we
find
is
Kafka.
The
way
this
works
is
we
set
up
the
caf-co
connector
there's
an
SDK.
You
can
use
to
write
more
of
these
integrations
and
then
on
your
function.
A
You
say
what
topic
it
cares
about
as
an
annotation
when
we
receive
traffic
it
links
the
two
together
and
invokes
it
via
that
gateway.
So,
if
that
same
scaling
and
unified
experience
for
anything
that
you
do
now,
we'll
just
touch
briefly
on
your
architecture
and
what
you've
put
together.
If
you
can
tell
us
how
yeah.
B
B
We
start
deploying
functions,
we
are
the
coast,
routing
or
monolithic
architecture,
and
we
are
going
straight
from
monolithic
to
function.
We
are
skipping
the
microservice
step,
the
microcell
layer,
and
one
thing
I
want
to
quickly
remark
here-
is
that
we
are
interacting
with
a
lot
of
languages,
a
lot
of
procedures.
What
we
had
we
now
now
already
have
runny
like
Pete
or
PHP
homepage,
or
many
many
processes.
Many
projects
writing
in
Java
with
a
gene
axis
but
I'm.
B
Even
we
have
our
PG
procedures
on
or
IBM
AI
series
that
we
consume
from
or
from
all
functions,
and
we
are
using
a
lot
of
the
ecosystem
with
a
cloud
native
ecosystem
like
Prometheus,
not
only
for
for
openpaas
out
of
scaling,
but
we
are
monitoring.
We
were
using
Prometheus
for
monitoring
or
our
own
infrastructure.
We
are
using
a
fauna
for
visibility,
but
also
we
are
firing
alerts
to
some
BOTS.
So
we
can
note
that
we
certain
functions
are
going
to
collide
course:
Canada
we
are
using
fluent
or
logger
aggregation
logging,
a
greater.
A
So
thanks
for
that,
you
know,
I
think
if
you
want
to
know
more
about
this
particular
architecture
come
and
ask
us
at
the
end
running
a
little
bit
tight
on
time
now.
So
these
are
some
of
our
top
end
users,
some
of
the
biggest
ones,
the
ones
that
are
allowed
to
say
that
they
use
this.
What
was
it
like
actually
going
to
your
boss
and
saying
can
can
we
get
listed
on
this
open
source
project
yeah.
B
B
A
B
As
one
of
the
most
interesting
project
right
now,
what
you
have
been
cursèd
is
is
project
when
we,
when
you
can
deposit
your
check,
just
skin
a
picture
picture
it
and
all
that
will
be
a
hundred
bit
by
open
functions.
We
also
have
a
lot
of
credit
scoring
models,
machine
learning,
models
that
we
are
moving
to
open
pass
and
we
also
are
trying
to
find
a
way
to
move
or
genexus
coded
code
generated
procedures
and
projects
to
become
them
functions.
That's.
A
Very
popular
in
Latin,
America,
cogeneration
and
we've
just
got
a
couple
of
minutes.
Left
I
want
to
tell
you
about
one
more
thing
before
we
finished
recording
and
then
we'll
take
some
questions
afterwards.
So
we've
started
doing
something
by
listening
to
the
users
and
seeing
where
are
the
pain
points?
What
do
you
actually
want
on
top
of
open
fires,
but
it
turns
out
people
want
and
managed
experience
that
is
multi-user
gives
you
a
personalized
and
has
C
ICD
built-in.
A
Now
we've
done
that
through
something
that
you
might
have
heard
called
guitars
and
you
get
your
feedback
through
github.
So
as
soon
as
you
do,
a
push
to
a
private
or
public
and
github
or
get
lab
repo
will
take
the
event
process.
It
build
and
package
your
function,
deploy
it
in
your
cluster
and
give
you
feedback.
You
then
get
a
nice
personalized
dashboard
that
looks
like
this
and
you
can
get
additional
data
like
what
Patricia
likes
the
statistics
find
out.
A
Is
it
breaking
and
so
really
to
wrap
up,
because
we
could
probably
talk
all
night
about
this.
Our
mission
and
open
fast
community
is
to
make
service
functions.
Simple.
That's
the
core
of
the
message
we
want
to
be
leading
in
open
source
service
like
we
have
been
doing
for
the
last
12
months,
I
mean
we
have
almost
16,000
github
stars,
and
it's
really
important
for
us
to
provide
the
best
faz
experience
possible.
Well,
not
overloading
you
with
technology.
A
Open
fast
cloud
is
going
to
allow
us
to
enable
teams
and
solve
problems
as
people
adopt
this
in
the
enterprise,
and
they
don't
want
to
custom
craft
their
own
solution,
but
they
wants
the
convenience
of
having
something
that's
ready
and
that
supports
them
now.
One
thing
that
I
think
is
particularly
interesting,
as
we
saw
this
in
the
keynote
with
Kelsey
is
the
migration
paths,
so
we
want
to
make
it
really
easy
for
you
to
get
from
lambda
to
open
fires
or
the
other
way
around.
A
We
heard
recently
that
the
founder
of
Q
bliss
has
moved
on
to
other
things
and
that
project
may
be
put
into
maintenance
mode.
So
again
we
want
to
help
the
whole
community
by
giving
them
an
easy
way
to
migrate
on
to
open,
fast
and
other
platforms.
So
thank
you
so
much
for
listening,
I
hope.
You've
enjoyed
the
talk,
we'll
be
here
at
the
end,
and
we've
got
one
more
thing
for
you,
which
is
three
ways
that
you
can
get
involved.
A
So
if
you
want
to
take
a
photo
of
this
I'll
leave
it
up,
you
can
get
connected
on
slack
patricia
is
there
is
one
of
the
most
active
members?
Thank
you
for
that.
You
can
learn
with
12
practical
labs,
how
to
build
functions
and
see
if
this
is
for
you,
and
the
last
thing
you
could
do
is,
but
if
you're
getting
into
the
hype-
and
you
want
to
push
things
a
bit
further-
you
can
make
use
of
ISTE.
Oh
we've
open
fires
and
walk
through
the
labs
that
we've
built
there.
Okay,.