►
From YouTube: Microsoft Graph connectors updates and Postman samples
Description
In this 18-minute developer focused demo, Brian T. Jackett shows how to use Microsoft Graph Connectors to get and surface external data in native Microsoft 365 experiences like search - GA, intelligent discovery, compliance, Viva Topics, and context IQ – in development, and People – in Preview. Recent platform updates include Schema Updates API, new AAD permissions and simplified ACL. Upcoming work described, see Microsoft Graph connector search demo and finally step through how and why build a connector using Postman. This PnP Community demo is taken from the weekly Microsoft 365 Platform Community call recorded on April 19, 2022.
Demo Presenter
• Brian T. Jackett (Microsoft) | @BrianTJackett
Supporting materials
• Documentation - Microsoft Graph connectors overview | aka.ms/graphconnectorsapi
• Documentation - Use Postman with the Microsoft Graph connectors API | https://docs.microsoft.com/graph/connecting-external-content-connectors-api-postman
• Demo - Latest on Microsoft Graph connectors SDK – Rajdeep Chandra (Microsoft) - https://youtu.be/rRLRXWjY11k
Learn more
• Microsoft 365 Unified Sample gallery - https://aka.ms/m365/samples
• Microsoft 365 Platform Community in YouTube - https://aka.ms/m365/videos
• Microsoft 365 Platform Community - http://aka.ms/m365pnp
A
So
for
those
of
you
who
are
unaware,
microsoft,
graph
connectors
is
really
about
being
able
to
take
external
data
and
have
that
data
surfacing
in
a
number
of
different
native
microsoft.
365
experiences
there's
a
few
of
them
that
are
already
ga,
including
things
like
search
and
we're
working
on
in
development,
as
well
as
in
preview.
A
A
couple
of
other
experiences
so
search
being
great
for
end
users
to
be
able
to
actually
go
out
and
search
for
that
content,
but
there's
other
areas
like
intelligent
discovery
where
we
can
actually
show
content
to
you
based
on
different
activities,
things
that
are
going
on
in
the
system.
So
knowing
like
what
are
recommendations
for
you,
what
are
recently
accessed
files
in
the
past?
That
was
all
content
that
was
just
inside
of
microsoft,
365,
but
now
we're
actually
opening
the
doors
for
that
to
be
also
from
external
content.
A
Well,
there's
actually
a
lot
of
different
data
sources
for
this.
It
might
be
a
on-premise
database.
It
might
be
a
line
of
business
system
that
you've
created
might
be
a
third-party
sas
service.
Maybe
a
file
share
sitting
on
some
kind
of
server,
and
so
the
idea
is
that,
having
that
content
being
able
to
integrate
that
directly
into
microsoft
graph,
then
we
can
actually
be
able
to
surface
that
in
those
different
kind
of
experiences.
A
So
if
you
have
something
come
from,
let's
say
a
set
of
research
articles
that
you've
got
in
some
kind
of
third-party
system.
Maybe
you've
got
content
again
in
an
on-prem
database
or
some
of
the
kind
of
repository
we
can
actually
integrate
that
directly
onto
the
all
tab
and
show
that
content
in
there.
It
is
fully
up
to
you,
in
terms
of
the
as
the
connector
author
being
able
to
modify
the
lookout
and
feel
for
this,
so
we're
using
adaptive
cards
to
actually
go
and
render
that
content
out
there.
A
So
if
you
want
to
scope
your
content
to
show
directly
inside
of
that
custom
vertical
out
there
another
option
for
you
to
be
able
to
show
that
content
in
there
and
again
now
we're
actually
going
beyond
just
that
little
small
cluster
of
results,
and
now
you
can
actually
get
the
full
set
of
them.
So
you'll
do
some
things
like
filtering:
be
able
to
actually
get
a
little
bit
deeper
into
that
content.
A
So
as
long
as
you've
properly
defined
your
schema
for
that
graph
connector,
we
actually
go
ahead
and
actually
use
some
of
that
metadata
properties
to
further
refine
that
content.
That's
coming
through.
Maybe
you've
got
what's
the
source
for
this,
maybe
the
file
type
for
this,
the
extension-
maybe
some
kind
of
date
filters
lots
of
ways
to
go,
to
interact
with
that
content
and
then
be
able
to
show
it
out
there
on
the
search
results
page
now,
a
couple
of
recent
releases
that
I
do
want
to
cover
for
folks.
A
A
The
first
thing
to
mention
is:
we
do
now
actually
support
a
limited
set
of
schema
updates
in
the
past
when
you
defined
your
connector
you'd
specify
that
schema
that
tells
us
hey.
Here's
all
the
metadata
properties
that
I'm
looking
to
update
for
this.
So
this
is
going
to
be
the
title-
the
modified
date,
the
author
for
this,
the
actual
content
of
the
things
that
are
out
there
in
the
past.
If
you
want
to
add
a
new
property,
if
you
wanted
to
add
a
new
semantic
label
to
that
or
an
alias
or
similar
kind
of
things.
A
Unfortunately,
you
have
to
tear
down
and
recreate
the
entire
connection
out
there.
These
days,
we
actually
do
support
some
limited
number
of
capabilities
in
terms
of
adding
new
properties
out
there,
adding
new
labels
modifying
the
aliases
that
are
out
there.
So
there's
a
few
of
these
that
are
supported
in
kind
of
a
initial
release.
A
For
this,
we
are
looking
at
additional
different
types
of
scenarios
for
this,
and
now,
when
you
think
about
this,
in
terms
of
what
kind
of
things
might
require
re-ingesting
content,
there
are
a
few
scenarios
that
that
does
be
the
case
again.
If
I'm
adding
new
properties
onto
there,
I
need
to
basically
re-index
that
content.
So
I
have
the
latest
and
greatest
out
there
to
show
in
search
results
to
show
the
other
experiences
out
there.
So
in
our
docs
we
do
call
out
what
are
those
scenarios
and
which
ones
do
require.
A
Re-Ingestion
just
note
to
everybody
that,
yes,
there
are
some
new
support
for
scenarios
like
that.
Additionally,
the
permissions
that
we
use
for
oauth
authentication
in
the
past,
they
were
a
little
bit
limited.
We
only
had
essentially
hey,
go
and
write
to
all
connections
out
there
or
write
to
the
ones
that
I
own,
but
that
was
using
application
permissions.
Basically,
those
unintended
kind
of
service
account
level
access
into
this.
A
We
didn't
really
have
as
much
support
for
delegated
permissions
or
for
as
fine
grain
of
permissions
through
talking
with
a
lot
of
our
partners
and
a
lot
of
our
customers.
We
realized
hey.
There
are
a
lot
of
other
scenarios
that
we
do
need
to
support,
so
we
worked
on
adding
that
support
into
there
and
we're
actually
very
very
close
to
publishing
the
docs
updates
to
this,
but
they're
actually
live
in
the
platform.
A
So,
if
you'd
like
to
go
out
and
test
these
things
out,
we
do
have
support
for
additional
properties
out
there
stay
tuned
for
the
docs
updates.
That
should
be
coming
very
very
shortly
and,
lastly,
for
our
access
control
list,
how
we
actually
secure
the
content
out
there.
Yes,
every
single
item
that
you
push
into
a
connector,
these
can
be
permissioned.
So
if
you
have
a
azure
id
group,
if
you
have
an
azure
id
user
or
we
even
support
things
called
external
groups,
you
can
kind
of
define
those.
A
Also,
a
quick
note
that
we
do
have
some
updates
in
terms
of
microsoft,
built
connectors
that
are
out
there
there's
a
couple
of
different
sources
from
jira
to
confluence,
to
servicenow
and
so
forth.
Those
are
actually
been
released
to
ga
within
the
last
couple
of
months,
so
great
to
see
more
platforms,
more
support
and
additional
things
are
being
supported
with
those
microsoft
built
connectors.
A
A
In
the
past
we
only
supported
700
000
items
per
connection
and
when
you
combine
that,
together
with
the
ten
connections
allowed
for
the
entire
tenant,
that
meant
that
you
only
had
seven
million
items
for
the
entire
tenant
there
very
soon
we're
actually
in
the
midst
of
rolling
this
out.
So
this
will
be
here
pretty
pretty
shortly.
Some
of
you
may
may
have
already
seen
this
others.
It's
coming
to
your
attendance
very
quickly.
A
Here
are
a
couple
of
our
connections
that
have
that
increased
scale
up
there,
so
you
can
already
see
that
they
have
again
the
limit
for
5
million
items
and
they're,
storing
very
close
to
that
in
terms
of
numbers
that
are
in
there
and
then
pairing
that,
together
with
what
is
available
from
a
throughput
perspective.
Currently,
it
is
set
at
four
items
per
second
that
you
can
be
able
to
ingest
at
a
time
with
this
increased
scale
for
number
of
items.
We're
also
increasing
the
throughput.
A
So
we're
going
to
be
getting
up
to
20
documents
or
20
items
per
second
that
you
can
be
able
to
ingest.
So
a
lot
faster
for
us
to
be
able
to
fill
that
connection
up
as
well
as
do
things
like
reindexing
or
incremental
crawls.
Things
like
that
to
go
and
update
so
great
western
c
additional
scale.
Being
able
to
go
with
that
and
then
lastly,
we
did
actually
have
a
previous
community
call
presentation
on
this.
This
is
for
our
graph
connectors
sdk.
A
If
you
are
building
your
own
line
of
business
or
kind
of
your
own
custom
connector
out
there,
it
is
possible
to
hit
use
this
sdk
to
actually
go
out
and
kind
of
ease.
The
on-ramp,
for
this
basically
provides
a
number
of
the
kind
of
behind-the-scenes
plumbing
some
of
the
different
kind
of
sample
code
snippets
that
you'll
need
to
be
able
to
go
and
implement
these
things.
A
Please
take
a
look
at
the
previous
community
call
we'll
make
sure
that
that's
in
the
the
playlist
out
there,
so
you
can
take
a
look
and
see
how
to
build
connectors
with
the
sdk,
with
this
I'll
jump
over
to
our
docs.
So
I
can
actually
show
you
a
little
bit
about
how
to
be
able
to
build
your
connectors
and
today,
we'll
focus
a
little
bit
on
using
postman
to
go
and
do
that.
So
the
first
thing
I
want
to
do
is
highlight
that
we
can
go
out
to
our
documentation
for
this.
A
A
As
I
mentioned
a
little
bit
earlier,
that's
been
able
to
do
like
service
accounts
or
kind
of
unattended
queries
against
things,
and
so
for
with
graph
connectors.
There
are
a
few
key
scenarios
that
you
have
to
use
application
permissions
for
this,
and
so
because
of
this,
we
do
have
a
sample
set
of
different
apis
and
requests
that
you
can
make,
and
we
built
all
those
requests
for
graph
connectors
and
we
bundle
them
together
with
the
microsoft
graph
set
of
apis
and
samples
that
are
out
there.
A
So
this
article
out
here
on
this
page,
this
will
walk
you
through
a
little
bit
of
how
do
I
actually
go
and
clone
kind
of
fork
that
repo
out
there?
How
do
I
import
that
into
postman
be
able
to
work
with
the
different
apis
that
are
out
there?
It'll
show
you
a
little
bit
about
actually
creating
your
own
azure
id
application.
A
I
can
go
under
application,
which
is
going
to
be
where
I've
got
a
lot
of
my
different
sample
queries
going
against
graph,
connectors
and
you'll
see.
We've
got
this
split
out
in
a
couple
of
different
areas,
there's
some
of
the
initial
things
for
creating
your
connection,
actually
setting
it
up.
If
you
need
to
register
your
schema,
you
can
go
ahead
and
get
the
schema.
You
can
go
and
get
the
connection
all
those
kind
of
great
things
for
doing
just
the
administration
up
front
for
things.
A
We
also
have
for
doing
content
sync
now.
This
is
my
own
custom
postman
collection
out
there,
so
I've
got
a
couple
of
additional
things
that
you
may
not
see
in
your
own.
This
is
just
for
me
from
doing
demo
purposes,
but
I
can
go
ahead
and
do
the
upserts
there's
either
an
update
or
an
insert
for
a
number
of
different
items
out
there.
I
can
go
and
get
items
I
can
do
a
partial
update.
I
can
delete
items
so
great
ways
to
go
and
interact
with
that
content.
That's
out
there
and
as
well.
A
If
you
want
to
do
things
for
identity
sync,
as
I
mentioned
before,
you
can
set
those
apples,
the
access
control
lists
out
there
and
for
this
sometimes
you
actually
need
to
go
and
do
those
external
groups.
So
here's
a
great
way
you
can
actually
test
those
things
out,
be
able
to
go
and
create
that
external
group
I'll
be
able
to
populate
that
with
azure
id
users
and
other
kind
of
ways
to
integrate
with
that
identity
areas.
A
Now
I
see
a
great
question
from
the
chat,
which
is:
is
this
something
akin
to
bcs
for
those
of
you
who
are
not
familiar?
That
was
the
business
connectivity
services.
Part
of
sharepoint
was
that
2013?
I
believe
I
could
say
this
is
similar
to
that,
but
it
is
not
going
to
be
functionally
equivalent
like
one
for
one
on
that.
A
Bcs
was
really
about
being
able
to
connect
to
an
external
data
source
and
display
that
as
a
sharepoint
list
inside
of
sharepoint,
2013,
sharepoint,
online
and
other
areas
now
with
graph
connectors,
we
can
ingest
content
in,
but
it's
really
not
going
to
be
about
doing
that
real
time
in
the
ui
updates
to
that.
Instead,
it's
going
to
be
more
of
displaying
this
in
a
number
of
different
capacities,
so
think
of
it
more
like
a
one-way
ingestion.
A
A
The
other
key
difference,
though,
is
also
that
inside
of
m365
we're
showing
in
a
lot
of
different
areas.
Not
only
do
we
have
the
same
side
of
search
hosts
so
again,
sharepoint
puffs.com,
bing,
we're
also
looking
at
support
for
areas
like
teams
as
well
as
windows,
desktop
so
areas
that
you
know,
bcs
didn't
really
play
in
previously,
as
well
as
integrating
with
new
platforms,
again
viva
topics
in
terms
of
the
offs.com
recommendations
and
a
couple
of
other
areas.
A
So
there
are
some
other
ways
to
be
able
to
integrate
with
us
that
go
outside
of
what
pcs
supported
now
swinging
back
to
the
demo
on
here.
So
here
I've
got
this
under
application.
I'm
actually
going
to
go
ahead
and
get
my
access
token.
For
this
now
I
happen
to
have
a
whole
bunch
of
my
secrets,
for
my
environment
already
set
up
with
this.
So
when
we
think
about
you,
know
the
client
id
the
secret
to
actually
going
authenticate
for
this.
I've
already
plugged
these
into
my
environment.
A
So
when
I
go
ahead
and
do
the
get
new
access
token
go
go
ahead
and
proceed,
I'm
going
to
go
ahead
and
use
this
token.
It's
actually
going
to
save
it
into
a
variable
for
me.
So
I
can
actually
use
this
in
further
requests
and
when
I
come
down
to
like
my
get
all
collections
you
can
see
here
is
the
graph
request
that
we're
going
to
be
firing
off.
Craftymyself.Com,
v1,
external
slash
connections
go
ahead
and
send
this
request,
and
I
want
to
take
a
look
at
the
bottom
down
here.
Here.
A
Is
my
json
payload
return
for
this?
You
can
see.
I've
got
two
connections
in
the
sample
environment
in
here,
and
if
I
go
over
to
my
search
administration
center,
we
can
go
over
to
data
sources.
You
can
see
that
those
match
up
with
the
two
connectors
that
I've
created
custom
in
this
environment.
There
we
go
contessa,
btj
files
and
btj
test,
and
I
can
go
and
go
ahead
and
take
a
look
at
some
other
things
like
get
the
schema
for
this
go
ahead
and
just
grab
the
first
schema.
A
That's
out
there
again
great
way
to
kind
of
validate
what
are
all
the
properties
that
I'm
have
defined
for
this
things
like
the
id
for
it,
the
title:
what
is
the
extension
for
the
item
that
I'm
pulling
in?
Was
it
who's
it
created
by
what
was
it
created
date,
time,
last,
modified
and
so
forth?
So
I
can
actually
go
and
validate
what
is
that
schema
for
this
connection
out
there?
A
If
I
want
to
go
and
update
items
out
there,
I
can
also
do
that
with
our
simple
puts
on
these,
and
with
these
our
payload
can
be
either
the
full
item
itself.
If
you
want
to
spotify
sorry
specify
every
single
property
that
exists
on
that
source
system,
or
if
you
only
want
to
do
a
partial
update,
you
can
do
that
as
well.
It's
a
great
way
to
have
kind
of
flexibility
in
terms
of
hey.
A
So
if
I
go
ahead
and
pull
back
just
a
simple
item-
that's
out
there
already
I've
got
here.
Let's
say
my
item
three
and
again,
I'm
using
the
same
authorization
token
for
all
these
different
requests.
So
it's
kind
of
a
get
that
token
once
be
able
to
come
back
and
run
multiple
queries
that
are
out
there.
But
here
you
can
see
one
of
my
items
again
I
can
see
and
who
is
access
for
this
item?
Three
actually
has
everyone
in
the
tenant
can
go
and
access
this?
A
So
if
you
have
a
third-party
search
portal,
if
you
have
an
internet,
if
you
have
some
other
kind
of
thing,
that
you're
trying
to
show
the
content
in
great
way
to
be
able
to
leverage
microsoft
graph
apis
according
to
the
search
apis,
pull
that
content
back
out
and
actually
render
this
in
multiple
locations
so
great
way
to
see
where
we
can
surface
this
content
in
multiple
different
locations
and
just
show
you.
This
isn't
smoke
and
mirrors,
we'll
actually
pull
up.
Office.Com.
A
And
if
we
go
over
here
to
that
btj
files,
you
can
see
that
yes,
there's
like
that
sample
item
three
and
again,
I
have
the
permissions
so
that
all
users
can
see
this
content,
but
because
this
user
isn't
actually
in
all
the
other
groups
that
are
out
there,
this
user's
not
seeing
items
one
two
and
five
so
great
way
to
just
kind
of
see
the
desk.
This
is
not
smoke
and
mirrors.