It's always great
to be back at my old
stomping ground.
Shout out to Brooklyn.
okay, Brooklyn,
I see you.
To prepare for
this, I did a little
retail research and
went to the Canada
Goose store, and I
was not alone. It
looks like our
amazing demo drivers,
Miles and Gus
also did too.
And listen, when I went
to that store, I can
attest to the fact
that Canada Goose is a
global leader in premier
performance luxury.
Now, even the best
of brands struggle
with how to provide
that premier
service across
disconnected channels.
Now, before Agent
Force, providing that
high-touch service
required their teams
to jump frantically
from screen to
screen, piecing
together data, all the
while the customer
waiting. But today,
Canada Goose has
turned that customer
journey into a
continual conversation.
By automating over
89% of their routine
tasks, it's freed
up their team's time
to go ahead and focus
on that high-touch
service that drives
customer loyalty.
Now, I know many of
you are here in this
audience, and you may
have teams that are
feeling that same pain
point of providing that
luxury experience across
disconnected data.
But what if, what if
we could take away
the time from them
managing handoffs
and running around
getting together data
and focused entirely
on the customer?
Well, I'm going
to show you a demo
that's not a what
if, but a what is.
And it's going
to show you how
we can make
this a reality.
We're going to
start with an
amazing voice
first experience.
And then we're
going to jump into
the power, the
full power of the
Salesforce platform
and see how Data360
really powers
that experience.
And then we're going
to look at how do
we bake in trust
and control? How do
we create these
amazing experiences?
We're going to pull
back the curtains
a little bit. Are
you all up for this?
All right. So
let's get into it.
So here we are at the
Canada Goose website.
And Fiona,
because Salesforce
is headless, has options.
She can either choose,
you know, messaging,
voice, or even
online. She chooses to
call. And because
this is built on the
world's number
one CRM, the agent
already has full
context of the customer.
I'd like to
make a return.
Got it. I have
your order here.
I also have your
contact details.
Would you like me
to send your return
label to your email
or through WhatsApp?
And I'm looking
for something
a little lighter
for spring.
Do you have any
recommendations?
Done. I've just sent
it to your WhatsApp.
For personalized
recommendations, I'll
connect you with a
Canada goose-style expert
who can help you find
the right option.
Now, that was perfect.
Thank you so much,
Fiona. And, you know,
Patrick was right.
Winter's over, so I
need a little lighter
jacket as well. That
was amazing. And
let's unpack what
we just experienced.
When Fiona
called in, they
immediately recognized
her. There was
no identification
or repetition.
And the agent didn't
just answer the
questions. It took
action, processing
the return and sending
the label to her
preferred channel,
which in this case
was WhatsApp. Now,
what also is really
interesting is what
you didn't see.
Agent Force
was behind the
scenes working
with the CRM,
pulling in data
from data 360, and
analyzing the voice
calls for intent
signals and rendering
it right here
in the Agent Force
command center.
But now, let's
jump into the
perspective of
the stylist.
I also want you to
note when Fiona had
a question about
styling is seamlessly
brought in an agent,
a stylist here. So
here we're looking at
Agent Force Command
Center. Now this
is the full power
of the platform.
Here this is not just
some customer contact
center bolted on
and stitched together
with some AI tool.
What you're seeing here
is voice, CRM data,
AI agents native to
the platform. So the
stylus has everything
they need right at their
fingertips to go ahead
and help the customer.
And because this is
powered by Data360,
the stylus has the
full context of the
customer, not just
the service incident.
Here they can see
their past purchases,
their preferences,
and every engagement
across the touch points.
And right alongside
the stylus, Agent
Force is going ahead
and serving up
recommendations so that the
stylus can focus
totally on the customer
and not playing catch
-up or searching.
Most AI systems,
they can answer
general questions.
I'll give them that.
But very few
can carry this
context, this
level of context,
into every customer
interaction.
Now, in my role,
I get to work
with our FTEs,
our Ford Deploy
engineers, shout
out to them.
Okay, they're in the back
because they're
working hard.
Now, here's the
thing. We get to work
with our customers
in building out our
roles and responsibility
of the agents.
And as you heard
Patrick say earlier,
is now exposed as an API,
I know that's a
lot of three-letter
acronyms, Leah, but this
is what democratizes
access to our
agents. It gives them
access to all the
goodness of the platform.
This is the essence
of headless.
Your customers aren't
tied to a certain
channel. They move, we
move, just like that.
We meet the customers
where they're at,
right? And here,
it allows us to
bring humans, agents,
and the platform
together in a
really powerful way.
So what I want to
do, if you all are
up for it, what I
want to do is pull
back the curtains
even a little bit more
and show you how
we were able to
create this magic.
Are you up for that?
So here we are in
Agent Force Builder.
This is where we design
and deploy agents.
Now, in here, there's
a lot of goodness.
We're going to go
ahead and expose one
of our sub-agents.
Now, this is where
you define the behavior
of your agents,
what they can, but
even more importantly,
what they cannot do.
And if you notice,
this is all done in
natural language.
Now, I want you
to remember when
Fiona called in, the
agent knew who she
was. And I know
you're thinking,
well, how did that
happen, Leah? Well,
I'm going to tell
you. Right here,
we hadn't got the
information. We got their
name, their past
orders, all of their
information right there.
So the model is reasoning
in real time, grounded
in customer data.
Now, also, if you
remember, when
Fiona started asking
style questions,
the agent was
like, whoa, let me
bring in a human
to handle that.
that's because we put
in a guardrail that
said right there when
when asking for styling
bring in a human that
my friends is not
a suggestion that is
a rule that cannot be
broken just think about
in financial services
or health care a wrong
answer isn't just
a bad customer
experience it can be a
compliance risk. And
we don't want those. So
here, we define it
once, and it's executed
everywhere. Built once,
activated everywhere.
Now, I want to pull
back the curtains just
a little further. You
guys open for that?
Yeah? Okay. You still
with me? All right.
So now what I want to
do, I'm going to show
you actually the code
behind this Agent 4
script, right? the
power of Agent Force
script. This code right
here is the code behind
the agent. Remember
the agent that Fiona
was engaging with?
Just nod your head.
Yeah, I remember that,
Leah. This is the code
that built that
agent. And now you're
thinking, okay, so
what? That's great. Most
people think, okay, a
developer wrote that.
Plot twist. No. This
was built by a coding
agent here in Agent
Force vibes. Now,
you could use Claude
Code, Codex, Cursor,
your choice. It
doesn't matter. It
all works in Agent
Force. But what you're
seeing is an agent
created by an agent.
I know, shut the
front door. This is
amazing. And what it
does is it lets you
go beyond the limits
of human capacity.
Now, don't get me
wrong. I love to
code. But you know
what I love better?
Watching my ideas come
to fruition faster.
That I love more.
But what you're
seeing is human
design, agent built.
That is an agentic
enterprise.