►
Description
How to get a Blockchain Project in Production - Marco Alarcon & Andrés Araya, Bolsa Comercio Santiago
At the Santiago Stock Exchange, we underwent a research project to be the first to implement a Blockchain based solution for the equity market in Chile.
In this session we will expose how we gestated this project since the design phase until how we started to deploy and run this in production, with all the problems that arose and how we solved them, using Hyperledger Fabric as the base framework.
We will expose the implemented architecture that allows today to operate this solution, and how this platform helps to reduce costs, improve trust between participants and enable new business models.
----
Hyperledger – Open Source Blockchain Technologies
Hyperledger is an open source community focused on developing a suite of stable frameworks, tools and libraries for enterprise-grade blockchain deployments.
It serves as a neutral home for various distributed ledger frameworks including Hyperledger Besu, Fabric, Sawtooth, Indy, as well as tools like Hyperledger Avalon, Cactus and libraries like Hyperledger Aries, Ursa. Learn more about Hyperledger projects: https://www.hyperledger.org/use
Case Studies: https://www.hyperledger.org/learn/case-studies
Training & Certification: https://www.hyperledger.org/learn/training
Tutorials: https://www.hyperledger.org/use/tutorials
Webinars: https://www.hyperledger.org/learn/webinars
Events: https://www.hyperledger.org/events
Vendor Directory: https://www.hyperledger.org/use/vendor-directory
Subscribe to the Hyperledger Newsletter: https://www.hyperledger.org/newsletter
Follow-us on Twitter @Hyperledger
Learn about Hyperledger Membership: https://www.hyperledger.org/about/join
#Hyperledger #Blockchain
A
Well,
welcome
I'm,
very
glad
to
be
here
in
the
first
IP
layer
called
Global
Forum.
My
name
is
Andy
today.
I
am
very
happy
that
I
won't
explain
what
is
blockchain
for
the
very
first
time.
The
last
two
years
today
is
a
very
technical
conference,
but
first
to
start
I
want
to
made
an
introduction
at
what
we
are
and
what
we
are
doing
in
in
South
America.
A
The
Santiago
exchange
is
the
mayor
exchange
in
Chile,
and
it's
an
hundred
and
twenty
years
old
company
we
are
in
South
America
is
the
chill
is
the
most
stable
country
in
Latin
America.
Economically
speaking,
we
are
very
open
economy.
We
have
more
than
25
international
trade
agreements
all
around
the
world
that
produce
a
very
competitive
economy.
We
are
open
an
open
ecosystem
for
start-up.
Our
core
government
introduced
a
program
for
startup
all
around
the
world,
many
developers
and
enterpreneurship
scum
to
our
country
that
fill
out
a
lot
of
interesting
resources
for
doing
new
things.
A
Our
country
is
also
recognized
to
produce
a
lot
of
copper.
More
than
30%
of
the
copper
around
the
world
is
produced
in
in
Chile.
It's
a
nice
place
to
visit.
We
have
the
driest
desert
in
the
north
of
Chile
and
a
lot
of
rain
in
the
South,
very
green
and
in
the
center
we
have
a
beautiful
undies.
You
can
go
to
the
skin
more
go
to
the
beach
in
the
afternoon.
So
a
nice
place
to
visit.
We
have
the
major
reserve
of
lithium.
A
A
More
than
3.5
billion
dollars
a
day,
trade
major
in
fixed
income
and
one
interesting
thing
that
we
are
mostly
a
technological
company,
more
than
60%
of
our
income
sources
come
from
products
and
services
that
we
sell
in
the
industry,
as
I
mention
before
we.
We
have
a
hundred
and
25
years
old
company
and
in
2011
what
was
we
create
international
agreement
with
Mexico,
Colombia,
Peru
and
Chilean
that
call
Miele
Alliance,
it's
an
andean
agreement
and
we
trade
in
real-time
with
Peru,
Colombia
and
Mexico
our
stocks.
In
real
time.
A
We
are
an
open
company,
a
public
company,
we
trade
in
the
stock
exchange
and,
for
example,
B
3
is
the
Brazilian
stock
exchange
has
10%
of
the
our
our
company
and
we
are
working
with
related
fabric
as
a
major
fairy-book
that
we
are
produced.
Software
related
to
blockchain
technology,
I
said
so
before.
Santiago
changes
and
major
technological
companies,
60%
of
the
income,
come
from
technological
products
and
services
that
we
provide
to
the
different
companies
specially
brokerage
houses
in
Chile.
A
A
If
something
happened
with
Apple
and
my
feeling
is
right
and
the
price
go
down,
then
I
buy
back
the
Apple
stock
in
the
market
and
give
it
back
to
Marco.
There
are
apples,
so
I
get
that
capital
gain
with
that
short
sales
that
how
its
work,
obviously
Marco,
to
lend
their
stock
I
have
to
sign
a
lot
of
contracts
for
collateral.
So
for
many
things
and
that
produce
friction.
A
back-office
frictions
that
allowed
that
this
market
is
is
small
in
Chile,
its
represent,
not
more
than
1%,
of
the
spot
mark
and
in
equities.
A
The
reason,
why
is
a
friction
back
office
services
for
what
we
have
to
do
and
that
appear
the
blockchain
opportunity
to
facilitate
that
kind
of
transactions?
When
you
start
with
our
product
services
in
blockchain,
you
have
to
select
very
well
what
kind
of
project
you
start
with
one
thing,
especially
in
Chile,
is
the
regulation.
We
are
a
regulated
company.
A
So
if
we
wait
for
the
regulator
change
the
rules
forget
about
the
law,
you
could
wait
three
years
for
a
start-up
project,
so
the
project
that
do
that
you
have
to
select
don't
have
to
change,
doesn't
have
to
change
any
rules
in
the
market
and
what
we
select.
This
kind
of
system
I
want
to
little
explain
in
business
side,
because
Marko
will
go
into
details
of
the
technical
size.
What
we
have
Santiago
change,
blockchain
network
looks
like
when
you
start
a
project
in
a
blockchain
space.
A
You
can
wait
that
every
single
company
that
form
your
ecosystem
adopt
their
own
system
to
the
you
new
technology.
So
we
have
three
use
cases:
the
the
blue,
the
first
brokerage
houses.
They
don't
are
intensive
in
this
market
short
sales,
so
they
don't
want
to
invest
anything
to
change
their
system.
So
we
provide,
through
our
terminal
network,
a
graphic
user
interface
about
a
web-based
application.
A
The
input
necessary
to
deal
with
short
sales
if
they
want
to
go
into.
So
they
do
nothing
around
this
modification
in
this
own
system
with
the
blockchain
technology.
The
second
case
is
the
blue
light,
the
second
brokerage
houses.
They
have
their
own
back
office
services
and
we
created
an
API
to
interact
from
their
own
back
office
services
with
our
own
infrastructure.
We
create
our
own
infrastructure,
on-premise
and
cloud
approximately
six,
because
this
kind
of
broker
brokered
they
don't
want
to
create
their
own
peer,
their
own
note
and
the
third
one.
A
The
green
one
broker
tells
us
they
want
to
go
into
the
technology
and
they
want
to
be
part
of
this
network
because
they
want
to
modify
their
own
back
office
services
and
introduce
this
new
database.
The
centralized
database
in
their
own
back
office
services,
and
they
do.
We
have
client
that
they
do
and
they
at
the
same
time
create
they
don't
know
their
own
peer.
Another
important
peer
in
this
network
is
obviously
Santa
our
exchange,
who
act
as
an
administrator.
A
We
are
created,
CA,
a
certificate
authority
and
we
issue
the
certificate
that
the
invitation
for
the
peers
and
institutional
to
work
with
this
network.
Another
important
peered
is
the
regulator
before
that.
We
launch
this
system
the
regulators.
If
they
want
to
analytics
across
the
the
the
short
sales
market,
they
have
to
request
information
to
the
brokers
and
institutions
collect
that
txt
files
and
after
a
week
they
do
the
analysis.
Now
they
are
doing
almost
in
real
time.
The
analysis
for
this
market,
another
important
period,
is
institutional
clients.
Pension
funds
usually
are
the
most
lending.
A
Institutions
are
allowed
to
lend
because
usually
they
buy
for
long,
so
they
allowed
to
do
to
lend
a
lot
of
shares
for
the
market
and
to
finalize
these
introductions
from
the
business
side
of
what
we
are
doing.
I
want
to
explain
a
use
case
when
you
start
dealing
with
short
sales
market.
A
client
have
to
call
to
the
brokerage
house
and
for
the
very
first
time,
sign
a
Master
Agreement,
so
in
the
old
world,
usually
a
client
called
to
their
broker
and
said:
hey
I,
to
start
dealing
with
short
sales.
A
A
There
are
nodes
or
our
web-based
application,
write
it
in
the
blockchain,
the
contract,
the
blockchain
system
send
to
the
client
directly
the
contract
launch,
but
the
smart
contract.
The
client
signed
the
emails
and
the
contract
who
is
recorded
in
the
blockchain
and
that's
it.
They
are
allowed
to
deal
with
short
sale
market
in
less
than
60
seconds.
That's
how
we
apply
in
one
case:
smart
contract,
our
obligation.
B
Thank
you
guys
for
making
clear
that
Chile
is
not
only
a
spicy
sauce
so
for
the
next
three
hours.
So
no,
we
don't
have
time
well.
For
the
remaining
time
before
lunch,
I
will
tell
you
a
story.
This
story
talks
about
how
we
built
the
blockchain
lending
repository
starting
from
scratch.
So
let's
go
and
check
this
ok,
we
started
over
there
in
January
2007
when
we
went
to
an
eye
to
the
IBM
King
garage
located
in
New
York
for
a
full
week
with
a
suitcase
full
of
probable
business
cases.
B
There
we
counted
with
the
expertise
of
both
technical
business
consultants
from
IBM
that
were
working
at
that
time
in
the
alleged
project.
What
was
the
idea
is
what
to
select
viable
business
case
that
we
could
use
as
a
pilot
to
build
a
minimal,
viable
project
to
select
this
particular
business
case.
We
employed
a
scoring
framework
that
was
used
at
IBM.
The
main
restriction,
as
under
said,
was
that
there
were
that
we
didn't
need
to
change
the
regulation.
Why?
Because
it
takes
a
lot
of
time,
sometimes
a
lot
of
money
to
do
that.
B
Okay,
then,
we
selected
in
this
case
the
short
sale
probe
discharge
same
as
the
problem
to
solve
and
move
to
blockchain
and
after
that
I
assembled,
a
small
team
of
very
capable
developers
experienced
it
in
end-to-end
technologies
and
in
conjunction
with
a
business
analyst
and
over
there
we
went
to
Baton
Rouge
Louisiana
in
the
u.s.
to
work
with
for
a
full
month
deep
dive
with
an
IBM
development
team.
B
B
It
was
a
proof
of
concept,
but
they
they
would
part
that
was
we
got
the
skills
necessary
to
keep
going
on
around
to
finish
this
product,
then
we
went
back
to
Chile
and
for
the
next
month
we
created
the
first
Network
and
eBay
basic
infrastructure
for
running
this
using
containers.
This
was
the
first
project
in
the
exchange
to
use
container
technologies
and
we
use
a
darker
white
docker
because
we
use
we
are
using
docker.
Also,
we
use
a
doggy
compost
for
local
envira
for
a
local
development
environment.
That
computer
is
my
development,
environment.
B
B
B
B
Okay
over
there,
you
can
see
the
back
end.
This
is
the
back
end
of
the
landing
repository.
This
is
a
blockchain
network
based
on
fabric
in
this
case,
as
under
smashing
it
we
are
the
owner
of
the
network
and
we
are
the
one
who
decide
who
participates
in
this
party
issuing
the
certificates.
We
are
running
the
certificate
authority.
We
create
and
distribute
the
certificates
and
also
we
develop
and
deployed
the
smart
contract
for
this
network.
We
are
running
peer
nodes
for
brokers
that
have
their
infrastructure
and
for
foreigner
purpose.
B
The
all
day
order
service
is
own
source.
We
are
using
Kafka
zookeeper
as
the
consensus
alberene
right
now.
Okay,
how
you
can
access
this
platform?
You
have
a
collection
of
REST
API
is
built
on
no
GS.
That
also
include
a
specific,
a
fabric,
clay,
a
client,
a
front-end
user,
API
integration,
API
for
other
system
like
the
exchange
air
P
for
brokerage
houses
and
also
at
a
schedule
that
is
basic,
sorry
on
Jenkins
CIE.
B
The
idea
donor
is
that
the
brokerage
house
can
use
this
solution
in
trifa
in
a
tree
fashion
way.
One
is
with
our
current
trading
platform
a
desktop
client,
the
other
one
is
that
the
brokers
can
integrate
their
back
office
system
using
the
REST
API
for
the
desktop
a
trading
platform
we
created
a
couple
of
HTML
view
is
modified,
some
of
the
pre-existing
UI's
to
interface
with
the
new
API.
B
And
then
the
solution
requires
also
the
displaying
and
signing
the
digital
documents
and
send
them
by
email.
So
we
had
to
create
all
the
required
infrastructure.
This
is
based
on
Jasper,
server
and
client
diamond
for
a
an
email
address
and
a
dedicated
API
for
all
document
management
function
and,
finally,
down
there.
We
have
the
participating
brokerage
house
that
is
directly
interconnected
with
our
fabric
network.
B
B
B
Okay,
we
have
here
an
eye
breed
schema
where
we
mix
a
couple
of
cloud.
Servers
in
this
case
is
a
Microsoft
Azure
and
our
traditional
data
center
back
in
Santiago.
We
have
two
data
center.
We
have
provided
in
just
one
of
them
in
GTD,
and
all
of
this
from
the
doctors
perspective
is
just
a
single
overlay
network,
independent
of
the
country
where
it's
located
so
I.
Think
that,
with
this
answer
in
some
question
that
I
heard
yesterday,
yes,
you
can
mix
even
countries
in
yourself
in
your
blocks,
blockchain
solution.
How
latency
affect
this?
B
Well,
it
depends
on
your
business
and
operational
requirements
here.
Just
for
that
we
are
running
in
each
doctor.
Node
many
containers,
each
containers
runs
at
least
one
pair
one
world
state
database.
That
is
coach
TB
and
we
run
all
the
order
service
in
the
cloud
because,
mainly
are
for
dissolution
all
most
service.
Most
servers
are
hosted
in
the
cloud,
so
we
wanted
to
optimize
latency
and
throughput
for
the
transactions
they
know.
The
peer
nodes
back
in
Chile
are
used
mainly
for
two
purposes.
One
is
to
do
queries.
B
B
B
We
are
running
this
in
a
high
availability
set
up
in
an
active/passive
cluster
because
is
running
with
postcode
sequence
that
has
to
run
in
an
active/passive
cluster.
That's
mainly
why
we
have
created
two
TRC,
a
schema
one
with
the
root
CA
that
is
always
off
well,
not
is
always
off
is
on
just
when
we
created
this
network,
and
then
we
shut
it
down.
B
B
This
is
for
security
purposes,
because
if
a
particular
certificate
for
I
don't
know
appear
an
application,
an
admin
application,
a
user
application-
is
compromised.
You
can
revoke
that
participant
particular
certificate
or
in
a
case
that
an
organization
had
a
security
breach
and
all
the
certificate
were
compromised.
We
just
revoke
the
intermediate
certificate
authority,
but
we
don't
compromise
the
world
network.
B
B
It's
important
to
mention
that
this
cluster
is
never
access
set
by
end
users
or
you
is
or
other
directly
by
systems
there
are.
This
is
mediated
by
the
front-end
api's
by
another
set
of
API.
Is
data
hosted
there,
so
those
are
living
in
a
separated
villain,
and
this
is
like
a
DMZ
atom.
It
the
emitter
side,
they
emit
a
DMZ,
and
in
this
case
well
it's
also
an
active
active
cluster.
B
B
How
we
do
the
sizing
we
took
some
dices,
we
troll
them.
That
was
it
now
in
for
the
Dockers
worm
here.
This
is
how
they
heavy-lift
work
is
done.
We
provided
in
a
sure
we
also
tested
for
Amazon
Web
service.
A
couple
of
weeks
ago
we
had,
we
did
a
internal
test
to
move
a
couple
of
peer
to
another
cloud
and
interconnect
to
clouds.
Also,
we
have
for
Linux
servers
in
the
middle
tier
available,
16
gigabyte
rams
for
veto
cpu
in
chile.
We
provided
in
our
currency
center
the
same
infrastructure.
B
The
only
difference
was
the
storage
for
the
middleware
and
the
CA.
The
CIA
do
almost
no
work,
its
sleep,
almost
all
the
time
and
the
middleware
since
it
runs
many
service.
Well,
we
had
to
a
provision
a
similar
tier
that
we
use
it
for
the
fabric
nodes
for
the
CA
we
scaled
it
down
and
also
for
the
front
ap
a
front-end
api's,
because
the
front-end
API
is
are
mainly,
they
do
proxy
functions
for
request
and
store
only
authentic
Asian
and
user
data.
B
B
First,
in
the
model
layer
we
have
the
block
champion
and
all
the
related
stuff,
so
we
created
smart
contracts
to
manage
all
the
master
service
agreement.
The
short
sells
the
landings
information
and
a
couple
of
general
data,
smart
contracts
for
utilitarian
things
like
stop
rising
trading
hours,
etc.
B
The
since
there
is
sensible
information
that
the
brokers
don't
want
to
share
between
them,
it's
their
trade
secrets.
We
have
to
create
a
multi-channel
schema
where
we
have
master
channel,
that
the
regulator
and
the
exchange
can
access
general
public
channel,
that
stores,
information
like
stock
prices
and
specific
channel
for
the
brokers
where
they
can
create
transaction
on
their
own
behalf,
and
the
stock
exchange
also
can
access
and
create
transactions.
B
Moving
up
between
the
layers
between
the
business
and
the
model
layer,
we
have
a
low,
a
low
level
set
of
api's
that
are
in
charge
of
all
the
functionality
of
the
peers.
The
channels
were
contract
in
general,
it's
an
abstraction
of
the
fabric
client
and
over
there
there
we
have
the
controller.
This
is
mainly
an
API
that
do
business-related
logic,
use
it
the
low-level
API
to
leverage
the
access
to
the
network
and
also
has
an
admin
controller
that
is
in
charge
to
administer
the
network.
B
The
only
thing
that
does
mainly
is
adding
the
security.
Finally,
we
have
the
UI
views,
that
is,
a
mix
between
html5
for
the
new
functionalities,
like
the
master
service
agreement
manager
and
magre
integrated
functionalities
like
the
landing
management.
Also
in
this
layer,
we
have
access
to
all
the
common
functionality
de
or
that
or
trading
desktop
client
can
provide.
B
B
We
learn
it
a
lot
of
things,
but
this
is
a
couple
of
lesson
that
I'd
like
that
you
take
back
home
first
when
we
were
choosing
in
the
face
in
the
face
0
before
starting
the
real
project,
you
have
to
pick
a
suitable
business
case
right.
Ok,
don't
try
to
do.
Did
this
alone
by
yourself,
get
somewhat
with
experience
sounds
obvious.
He
inside
is
20/20
right,
use
it
as
core
frame
a
scoring
framework
and
and
most
importantly,
choose
something
that
is
a
very
quick
win,
but
doesn't
require
regulatory
changes.
B
Then,
when
we
did
the
deep
dive,
what
we
learn
it?
Yes,
we
we
were
in
that
building.
First,
for
your
Human
Resources
prioritized
quality
over
quantity.
You
can
do
this
better
with
a
small
team
of
multifunctional
developers
and
business
analysts
be
agile,
use,
agile
methodologies
for
your
minimal,
viable
project,
get
business
aware
developers,
you
need
developers
that
understand
the
business
domain
and
you
need
functional
business.
Analysts
that
understand
what
blockchain
is,
because
they
don't
have
to
be
great
one-to-one
your
solution
to
blockchain.
They
have
to
reinvent
your
solution.
B
Ok,
this
is
not
very
exciting.
Is
the
integration?
What
we
learn
it,
what
we
do,
the
full
backend
integration?
First,
it's
obvious!
It
will
take
a
lot
of
time
more
that
valued
what
you
can
think
and
if
you
see
that
is
a
long-term
effort
plan
ahead
and
includes
activities
to
do
the
upgrades
between
fabric
versions.
B
They
say
that
fabric
is
retro
compatible,
it's
99%
correct,
but
there
is
always
something
that
gets
broken
or
changed
or
duplicated
between
well
fronts.
So,
plan
ahead
this,
and
also,
if
you
have
right
now,
are
existing
soya
or
micro
service
architecture.
That
is
a
very
good
thing
that
will
help
you
a
lot.
B
B
You
have
to
plan
ahead
a
trial
period,
a
trial
period,
and
you
have
to
engage
about
your
internal
and
external
users,
and
this
is
really
important,
include
stress
and
resistant
tests
that
simulates
high
concurrency,
because
the
developer
is
single
threaded
in
his
mind.
He
will
always
think
that
he
will
test
the
system
thinking
in
one
user,
using
the
UI
and
using
all
the
functionalities.
He
never
do
a
simulation
with
hundreds
of
users
testing
his
function
and
in
this
case,
in
a
distributed
system
it
will
explode.
B
We
have
a
couple
of
very
bad
night's
fixing
some
code
there.
Also,
if
you
are,
you
have
a
blockchain
ever
you
want.
You
have
a
community
other
participants,
so
their
interconnection
will
be
as
low
as
their
telecom
and
your
own
telecom
teams
put
together
because
they
always
talk
a
different
dialect
of
binary
and
so
get
your
other
exterior
external
stakeholder,
the
other
organization
and
get
a
common
schedule
together
to
do
the
rollout.
B
Ok,
we
have
the
system
in
production
and
well.
We
want
to
do
some
things
that
in
a
traditional
system,
we
can
do
easily
right.
We
can
we
can
we
want
to
stop
this.
You
cannot
do
that
is
in
production
and
is
not
going
to
stop.
We
try
to.
We
did
a
test
to
shut
down
50%
of
the
network
and
then
where
the
work
keeps
going
on
without
problem.
B
One
of
the
things
that
it's
quite
hilarious
is
from
operational
departments.
They
will
ask
you,
please
fix
me,
a
specific
role
in
the
database,
because
a
price
is
type
of
wrongly.
You
cannot
do
that
here.
You
have
to
go
to
the
smart
contracts
right
and
how
you
can
explain
to
the
operational
teams
or
the
audit
team
that
you
cannot
do
that
and
you
cannot
fix
the
price
in
couple
of
seconds.
B
You
have
to
do
a
new
release
to
add
a
functionality
that
allows
you
to
fix
that
specific
record,
because
your
smart
contact
enforce
new
business
rules
and
also
well.
This
system
will
be
as
solid
as
the
weakest
component,
because
you
have
many
moving
in
this
machine
and,
yes,
your
blocks
network
will
be
solid,
but
you
may
depend
of
external
or
another
system
that
are
not
as
solid
as
your
blockchain
network.
B
Finally,
to
find
out
to
finalize
this
a
couple
more
lesson
that
you
can
take
home:
this
is
a
paradigm
shift.
Right
is
a
distributed
system.
You
need
developers
and
analysts
that
sing
in
a
distributed
way.
The
ledger
setup
can
be
really
hard.
The
first
time
for
us
was
really
hard
because
we
started
with
fabric
0-6
beta
and
right
now,
it's
more
easy
to
do,
because
you
have
templates
for
cloud
providers,
and
you
have
already
built
up
a
files
for
docker
compose
and
queue
Burnitz
when
we
started
last
year,
there
was
almost
nothing
like
that.
B
You
need,
as
I
mentioned
before,
but
I
want
to
make
sure
that
again
you
need
a
while
skill
set
team.
You
need
developers
that
know
security,
network
business
analysts
that
knows
the
technology,
the
understand
computer
science.
Other
thing
important.
You
want
to
be
great,
an
existing
product
functionality.
You
cannot
do
that,
one
to
one,
you
cannot
migrate.
Your
workflow
exactly
like
is
right
now.
B
Innovate
think
something
new
like
the
process
that
we
did
for
signing
the
master
service
agreement.
Also
right
now
the
documentation
of
fabric
is
top-notch
when
we
started
is
well
as
really
close
to
zero,
so
we
have
to
go
directly
to
the
source
code.
Sometimes
right
now,
the
source
code
is
as
good
as
documentation
for
something
is
for
our
computer
scientists
is
better
to
read
the
source
code,
sometimes
that
the
documentation,
also
they
you
will
have
problems
that
in
your
current
system,
can
easily
solve
getting
that
the
local
machine
time
in
blockchain
solution.
B
You
cannot
do
that.
You
cannot
ask
system
dot
time
because
appear
null
will
tell
eval
way
in
Santiago
in
us.
We
will
have
another
value
in
Singapore
will
have
another
value
and
you
cannot
get
it
to
a
consensus,
so
stochastic
random
functions
time,
basic
solution,
non
deterministic
events.
You
have
to
eliminate
that
of
your
logic,
then
I
hate
that
names
MVCC
concurrent
problem.
B
You
have
to
model
thinking
that
in
a
single
block
you
can
only
update
a
single
key,
a
specific
key.
You
cannot
do
to
update
over
the
same
key
in
the
same
block.
We
had
to
learn
that
the
hard
way
and
we
have
to
refactor
a
wall
smart
contract
because
of
that
so
take
that
lesson,
because
it
will
solve
you,
many
many
headaches
and
finally,
well
with
this,
we
wanted
to
test
how
CouchDB
can
be
stressed,
how
how
can
coach
DB
can
be
used
for
all
relate
query
related
functionality.
B
We
found
that
it
has
some
limitations,
some
are
hard-coded
directly
in
public
and
some
are
specific
or
coach.
The
beef
moving
from
sequel,
server
technology
or
sequel
technology
to
document
oriented
technology,
so
you
have
to
be
creative
with
workarounds
compressed
data,
create
a
binary
index,
etc.
There
is
a
lot
of
thing
that
is
not
your
day
to
day
system.
B
B
A
D
B
It's
a
slow
and
expensive
data
base.
Now
we
went
first.
Is
we
wanted
to
test
the
technology
with
a
real
business
case
and
integrate
that
in
the
core
of
our
system?
We
wanted
to
test
if
this
technology
was
maturity.
Now,
with
this
project,
we
think
that
is
moderate
enough
to
solve
some
problems
in
the
financial
sector,
not
all
problems.
Let.
A
Me
up
something:
the
problem
in
the
financial
area
is
mostly
the
reconciliation
problem,
so
the
clients,
our
clients,
have
back
office
services
and,
in
the
use
case
that
we
have,
they
have
to
type
in
their
own
centralized
database
the
same
as
the
stock
exchange
and
that
need
a
lot
of
reconciliation,
usually
so
when
we
move
a
node
in
the
data
center
client.
That
is
the
case
of
this
use
case
that
we
produce
the
clients
if
the
major
bank
in
Chile
introduced
their
database
for
the
stock
exchange
in
their
own
system.
A
So
they
cut
a
lot
of
integration
that
they
have
to
recognize
information
during
the
day.
I
just
show
one
use
case
that
produced
that
four
days
of
complete
their
workflows,
move
to
60
seconds,
but
other
another's
smart
contract
produced
that
reduce
reconciliation
process
that
we
have
to
do
between
the
stock
exchange
and
the
brokerage
houses.
Usually
during
the
day
for
this
market,
we
started
with
this
product
that
is
slow
product
in
terms
of
amount
of
transactions.
But
now
we
want
to
move
another
use
case.
A
A
Are
they
on
premises
the
clients
on
the
prime
yeah?
We
have.
We
create
a
blockchain
as
a
service
infrastructure,
with
our
nodes
on
premise
and
annasher,
to
provide
a
service
for
the
brokerage
houses
that
don't
want
to
create
their
own
infrastructure,
but
is
an
option.
Another
option
is
create
your
own
nodes.
We
provide
a
service,
we
create
the
infrastructure
to
provide
service
for
monetize
a
way
to
monetize
our
infrastructure
to
provide
to
the
brokerage
houses
that
do
want
to
invest
in
their
own
infrastructure
is
decentralized
because.
D
B
A
A
By
their
own
here
and-
and
we
think
that
the
most
this
one
useful
case
yeah-
that
is
one
cavatelli-
it's
the
useful
scape,
because
they
trust
in
this
infrastructure
and
they
move
away
their
own
database
and
include
in
their
own
back
office
system.
This
new
the
centralized
database
there
and
avoid
reconciliation
process
got
it.
Thank
you
another
question.
Just.
F
Thank
you
for
presentation.
Do
you
develop
the
Shing
cut
with
a
cross
development
team
from
several
stakeholders?
It
was
the
question
of
trust
and
how
do
you
include
the
regulator
in
the
production
network?
How
do
you
convince
and
persuade
stakeholders
to
appear
for
a
question
of
costs
because
we
secure
the
network
and
do
you
work
out
participants
which
secure
with
the
cruiser
Network
sorry
a
lot
of
questions?
A
lot
of
questions.
A
You
have
it's
very
important
that
you
can
onboard
the
regulator.
Regulator
have
very
interesting
in
this
technology
and
has
a
mentioned
before
they
took
more
than
one
week
to
get
the
information
of
the
blockchain
and
this
kind
of
system
to
get
analyze
the
information,
and
they
recognize
that
we
appear
of
this
network.
They
could
made
that
analyze
information
in
almost
real-time
so
for
them
was
a
hit
to
connect
to
the
network.
So
it
was
very
if,
in
that
sense,
I
think
that
the
regulator
is
the
winner
in
this.
A
Built
this
solution
with
our
own
team,
as
a
marker
set
before
they
stay
for
one
month
in
Baton,
Rouge,
Louisiana
development,
in
conjunction
with
the
IBM
team.
We
understand
the
technology
with
them.
We
understand
the
patterns
and
they
went
back
to
Chile
to
continue
the
developing
of
the
network.
The
most
heart
problem
after
the
network
is
installed,
is
to
connect
with
the
current
system
with
the
legacy
system,
the
trading
system,
the
custody
system,
the
back
office,
services
of
Raqqa,
dialysis
and
creative
urge
to
connect
with
the
network.