WEBVTT

1b2a0816-5918-4c00-b5ee-a62b0313dca0-0
00:00:00.120 --> 00:00:04.238
The AI may propose, simulate an act,
but the purpose, and the context,

1b2a0816-5918-4c00-b5ee-a62b0313dca0-1
00:00:04.238 --> 00:00:08.762
and the value systems on which these
agents are built are still controlled by

1b2a0816-5918-4c00-b5ee-a62b0313dca0-2
00:00:08.762 --> 00:00:09.400
the humans.

4d14f9ff-f2f9-41d3-a9f5-7f47e616e254-0
00:00:09.680 --> 00:00:12.306
And that's how you create the right
balance and what we call the "AI

4d14f9ff-f2f9-41d3-a9f5-7f47e616e254-1
00:00:12.306 --> 00:00:12.840
future-built."

dbe6c450-20af-406e-9d0b-a47ecfee16df-0
00:00:13.200 --> 00:00:14.840
Japjit, thank you so much for being here.

4665e2a5-a86d-4914-9351-b39c05b01dba-0
00:00:14.920 --> 00:00:18.392
At a high level,
what does BCG mean by the idea of agentic

4665e2a5-a86d-4914-9351-b39c05b01dba-1
00:00:18.392 --> 00:00:19.040
enterprise?

398bd1d0-db67-4d94-884b-0d702ee2ab91-0
00:00:19.520 --> 00:00:21.360
Thanks for the question,
it's the theme of the hour.

cdd5d1ec-6436-4564-8c77-1beeb6d48805-0
00:00:21.640 --> 00:00:24.577
Look,
we think it's the next stage of digital

cdd5d1ec-6436-4564-8c77-1beeb6d48805-1
00:00:24.577 --> 00:00:27.579
maturity,
where companies are moving away from

cdd5d1ec-6436-4564-8c77-1beeb6d48805-2
00:00:27.579 --> 00:00:31.219
using independent,
isolated AI tools to having AI agents

cdd5d1ec-6436-4564-8c77-1beeb6d48805-3
00:00:31.219 --> 00:00:32.880
embedded in the workflows.

bf5811e3-14a3-435a-a2a0-d652ef05d9fb-0
00:00:33.320 --> 00:00:37.355
And those who are doing that are moving
away from pilots to at-scale workflows

bf5811e3-14a3-435a-a2a0-d652ef05d9fb-1
00:00:37.355 --> 00:00:39.960
being reimagined with AI agents at the
core of it.

bea45772-c1af-43fc-8195-477cbd0177dd-0
00:00:40.240 --> 00:00:41.960
We call these companies AI future-built.

9fc19e0b-db84-4a67-b2b3-66c7579265a5-0
00:00:42.360 --> 00:00:45.611
There are only few of them,
but those who are delivering these

9fc19e0b-db84-4a67-b2b3-66c7579265a5-1
00:00:45.611 --> 00:00:48.656
agentic interfaces for their clients,
for their customers,

9fc19e0b-db84-4a67-b2b3-66c7579265a5-2
00:00:48.656 --> 00:00:51.960
and for the internal employees are
delivering outsized returns.

2205c28f-31a9-445e-8bd0-819ae64d9faf-0
00:00:52.160 --> 00:00:55.760
So what really changes when companies
move toward this agentic model?

49765549-fe0c-440e-afda-89bfb80436e3-0
00:00:55.840 --> 00:00:58.803
Actually,
the shift is more architectural than

49765549-fe0c-440e-afda-89bfb80436e3-1
00:00:58.803 --> 00:00:59.560
algorithmic.

36d1c46f-c76f-4ff0-9300-5a68d0e8cbb7-0
00:00:59.560 --> 00:01:01.760
It's not about having another clever
model in the system.

7788c2f6-1dd8-4e33-a326-18481b6ad7fd-0
00:01:02.040 --> 00:01:05.997
It's about rewriting how business
decisions are made―enabling your agents

7788c2f6-1dd8-4e33-a326-18481b6ad7fd-1
00:01:05.997 --> 00:01:10.008
to work across different systems,
observe what's happening, learn from it,

7788c2f6-1dd8-4e33-a326-18481b6ad7fd-2
00:01:10.008 --> 00:01:14.073
act on their behalf is where the next
shift happens and takes away the drag

7788c2f6-1dd8-4e33-a326-18481b6ad7fd-3
00:01:14.073 --> 00:01:16.320
away from how the processes are run today.

34d4cb8e-da42-491d-9e47-87e3c07df96c-0
00:01:16.920 --> 00:01:20.760
And where do humans fit in as these
agents take on more autonomy?

a34b5fde-09bf-421b-8e50-43c3ba532024-0
00:01:21.400 --> 00:01:25.919
Look, that's the most important part.
In almost all successful deployments of

a34b5fde-09bf-421b-8e50-43c3ba532024-1
00:01:25.919 --> 00:01:27.600
agentic AI that we have seen,

5b69765d-14a0-47e8-bad4-95b1400605f1-0
00:01:27.920 --> 00:01:29.160
humans are always at the center.

7c861b9d-5f95-4079-af56-a5cab16cc3e2-0
00:01:29.560 --> 00:01:32.535
These systems are built for augmented
automation,

7c861b9d-5f95-4079-af56-a5cab16cc3e2-1
00:01:32.535 --> 00:01:34.440
not replacement of human beings.

45b5e0c2-4a86-401f-9f9b-ee36107c8059-0
00:01:35.080 --> 00:01:38.827
Think about the human creativity and
judgment and ethics being the supervisory

45b5e0c2-4a86-401f-9f9b-ee36107c8059-1
00:01:38.827 --> 00:01:41.200
layer on top of all which these agents
are built.

2f299a62-e705-44a1-aa22-e36c3f92b789-0
00:01:41.560 --> 00:01:44.724
So we're not replacing them,
but enabling them to create better

2f299a62-e705-44a1-aa22-e36c3f92b789-1
00:01:44.724 --> 00:01:45.120
systems.

813505d2-e684-4eb6-b7c1-379a55a1d6ee-0
00:01:45.440 --> 00:01:49.540
The AI may propose, simulate an act,
but the purpose, and the context,

813505d2-e684-4eb6-b7c1-379a55a1d6ee-1
00:01:49.540 --> 00:01:54.044
and the value systems on which these
agents are built are still controlled by

813505d2-e684-4eb6-b7c1-379a55a1d6ee-2
00:01:54.044 --> 00:01:54.680
the humans.

fe388d1e-5940-4525-b02b-bb73a2884866-0
00:01:55.000 --> 00:01:57.652
And that's how you create the right
balance and what we call the AI

fe388d1e-5940-4525-b02b-bb73a2884866-1
00:01:57.652 --> 00:01:58.160
future-built.

b68e4db8-9857-4bb8-85cb-ee83c1fcb6e8-0
00:01:58.440 --> 00:01:59.320
Thank you so much.

7648c11c-f4ea-4539-8346-0102143da2dd-0
00:01:59.400 --> 00:01:59.800
Thank you.