WEBVTT

00:00:00.083 --> 00:00:04.212
We're putting AI at the core,
and we're rebuilding around it.

00:00:04.212 --> 00:00:09.676
We're making sure that every one of our
processes is well understood, documented,

00:00:09.676 --> 00:00:11.594
and automated where possible.

00:00:12.053 --> 00:00:14.055
You said something a little while ago
that I really liked,

00:00:14.055 --> 00:00:15.098
and it's also on your website.

00:00:15.140 --> 00:00:17.267
You said,
"The adoption of AI is not just an

00:00:17.267 --> 00:00:19.561
upgrade,
it's the cornerstone of our operational

00:00:19.561 --> 00:00:20.103
evolution."

00:00:20.103 --> 00:00:22.105
So I'd love to get your perspective on a
couple of things.

00:00:22.105 --> 00:00:24.149
The first one is that that quote of yours,

00:00:24.149 --> 00:00:26.234
you know,
"AI is the cornerstone of Capita's

00:00:26.234 --> 00:00:26.735
evolution."

00:00:26.860 --> 00:00:29.863
What does that mean for you exactly,
sort of practically, day-to-day?

00:00:30.071 --> 00:00:32.699
Well,
it means that we are rebuilding the

00:00:32.699 --> 00:00:33.199
company.

00:00:33.908 --> 00:00:37.162
We are opening Capita as it was,

00:00:37.787 --> 00:00:41.958
we're putting AI at the core,
and we're rebuilding around it.

00:00:41.958 --> 00:00:47.422
We're making sure that every one of our
processes is well understood, documented,

00:00:47.422 --> 00:00:49.382
and automated where possible.

00:00:49.549 --> 00:00:55.138
We're making sure that all of our data is
available and is used safely and

00:00:55.138 --> 00:00:59.142
responsibly to train and to deploy
AI-based services.

00:00:59.309 --> 00:01:04.689
We're making sure that all of our people
are getting a certain degree of literacy,

00:01:04.689 --> 00:01:08.318
understanding,
and even proficiency to deploy, and use,

00:01:08.318 --> 00:01:09.152
and work with

00:01:09.152 --> 00:01:11.529
AI.
It's making sure all of our value

00:01:11.529 --> 00:01:16.618
propositions have an AI component to them,
and it's making sure that every one of

00:01:16.618 --> 00:01:21.289
our enabling functions inside the company
is looking at AI for efficiency,

00:01:21.289 --> 00:01:25.043
for a better service,
and to just be able to do new things.

00:01:25.376 --> 00:01:31.049
So it is the one thing and the one theme
that is present in every one of our

00:01:31.049 --> 00:01:32.092
conversations.

00:01:32.258 --> 00:01:35.595
What about the role of the sort of big
tech players and the hyperscalers in this

00:01:35.595 --> 00:01:35.929
journey?

00:01:35.929 --> 00:01:39.974
I think in my perspective,
you were very fast at involving them in a

00:01:39.974 --> 00:01:44.646
very fundamental way to how you work and
really engaging in deep partnerships.

00:01:44.646 --> 00:01:45.772
Can you comment a little on that?

00:01:45.939 --> 00:01:46.314
Sure.

00:01:46.397 --> 00:01:50.360
I think when you want to be so aggressive
with something like AI,

00:01:50.360 --> 00:01:55.365
some people might be tempted to say "I'm
going to create everything. I've got this.

00:01:55.365 --> 00:01:55.448
"

00:01:55.698 --> 00:01:59.410
The reality is that there is too much
innovation going on.

00:01:59.702 --> 00:02:04.624
There is this is a big lift for a lot of
our partners in the market,

00:02:04.624 --> 00:02:09.879
and it would just be irresponsible to try
to recreate what they're doing.

00:02:09.879 --> 00:02:11.464
And I don't think we stood a chance.

00:02:11.840 --> 00:02:15.802
So we accepted that the big innovation at
the systems level,

00:02:15.802 --> 00:02:20.014
at the applications level,
at the platform levels, in the cloud,

00:02:20.014 --> 00:02:25.562
at the API levels,  and at the interfaces,
what's going to be coming from these sort

00:02:25.562 --> 00:02:27.063
of hyperscalers, right?

00:02:27.105 --> 00:02:31.651
The . . . you know, the big four right,
in the application space--so the

00:02:31.651 --> 00:02:35.613
Salesforce, ServiceNow,
on the platform around AWS, Microsoft.

00:02:35.864 --> 00:02:39.659
And that what we needed to do
was to go very, very deep with them to

00:02:39.659 --> 00:02:43.204
to leverage always the latest innovations,

00:02:43.204 --> 00:02:47.500
the latest capabilities,
and help our customers orchestrate the

00:02:47.500 --> 00:02:52.213
services and the capabilities that are
offered by these hyperscalers.

00:02:52.714 --> 00:02:58.136
And to really translate our customers'
business requirements into a solution

00:02:58.136 --> 00:03:01.598
that is delivered with our hyperscaler
partners.

00:03:01.973 --> 00:03:06.144
And then have our role being the one that
orchestrates the solution by

00:03:06.144 --> 00:03:10.899
orchestrating our teams, our colleagues,
but also the different microservices or

00:03:10.899 --> 00:03:14.068
agentic capabilities that the
hyperscalers are there.

00:03:14.444 --> 00:03:18.990
So hyperscalers is the critical
ingredient to us delivering the strategy

00:03:18.990 --> 00:03:21.576
on,
on the speed and the innovation level

00:03:21.576 --> 00:03:22.660
that is required.

00:03:22.785 --> 00:03:26.331
I think we talked a lot the last years
around sort of orchestration as a

00:03:26.331 --> 00:03:29.125
differentiator,
the best possible orchestration for that

00:03:29.125 --> 00:03:31.336
very specific context of that client,
right?

00:03:31.586 --> 00:03:34.672
Can you talk a little bit also about the
human side of it,

00:03:34.672 --> 00:03:38.843
your learnings on how do you actually
drive value through AI on the human side,

00:03:38.843 --> 00:03:41.179
both in terms of, you know,
your employees,

00:03:41.179 --> 00:03:42.680
but also on your client side?

00:03:42.931 --> 00:03:46.643
A lot of what we do,
we do it for regulated industries, right?

00:03:47.060 --> 00:03:52.023
So there is a very clear line as to
what's an efficiency,

00:03:52.023 --> 00:03:58.279
where you can automate for efficiency,
and where you need to have a very

00:03:58.279 --> 00:04:03.576
qualified human making the final call as
a service provision.

00:04:03.952 --> 00:04:09.916
They are also customers who don't want to
be engaging exclusively with an AI agent,

00:04:09.916 --> 00:04:12.252
even though they're very capable.

00:04:12.585 --> 00:04:16.047
And there are situations where we don't
want that either.

00:04:16.464 --> 00:04:22.637
So for us to be able to orchestrate what
is done by a human agent and what's done

00:04:22.637 --> 00:04:27.183
by an AI automation or an agentic entity
is super critical.

00:04:27.809 --> 00:04:34.440
So the key for us is at the very,
very least, under the very, very minimum,

00:04:34.440 --> 00:04:39.404
everyone should be empowered and
augmented by AI, right?

00:04:39.404 --> 00:04:40.280
That is clear.

00:04:40.863 --> 00:04:44.534
At the other end,
there are some things that are 100%

00:04:44.534 --> 00:04:48.037
activities that could be easily delegated
to an AI,

00:04:48.037 --> 00:04:52.667
but the real magic is knowing how to
orchestrate what's in between.

00:04:53.376 --> 00:04:58.256
And that's where we're teaching our
colleagues to how to work with agents,

00:04:58.256 --> 00:05:00.341
how to delegate stuff to agents.

00:05:00.508 --> 00:05:05.680
We also letting some agents help us with
the orchestration of humans and agents,

00:05:05.680 --> 00:05:09.976
but the beauty here is we can actually do
this process by process,

00:05:09.976 --> 00:05:15.023
customer situation by customer situation,
regulatory environment by regulatory

00:05:15.023 --> 00:05:18.401
environment,
and just make sure that we get it right

00:05:18.401 --> 00:05:19.569
every single time.

00:05:20.111 --> 00:05:20.778
Fantastic.

00:05:20.945 --> 00:05:21.571
Thank you very much.

00:05:21.654 --> 00:05:23.781
Thank you for your time and for your
partnership.

00:05:24.282 --> 00:05:25.616
Thanks for all the help you've given us.

00:05:25.742 --> 00:05:26.117
Thank you.