1
00:00:00,000 --> 00:00:01,229
(gentle music)

2
00:00:01,229 --> 00:00:04,060
- GIA was founded in 1931.

3
00:00:04,060 --> 00:00:07,930
In early 1940s, GIA came up with standards

4
00:00:07,930 --> 00:00:09,930
called four C's for diamond grading

5
00:00:09,930 --> 00:00:12,870
to determine the value of the gem.

6
00:00:12,870 --> 00:00:17,870
Those forces are color,
clarity, carat and cut.

7
00:00:18,000 --> 00:00:21,260
It takes about 12 to 18 months
for somebody to get trained

8
00:00:21,260 --> 00:00:23,856
and be productive in
diamond grading process.

9
00:00:23,856 --> 00:00:27,810
A grader will look at
the natural inclusions

10
00:00:27,810 --> 00:00:29,800
within that diamond through the microscope

11
00:00:29,800 --> 00:00:34,328
and would plot them specifically
on a computer system

12
00:00:34,328 --> 00:00:37,140
to reflect off what they see visually.

13
00:00:37,140 --> 00:00:39,010
There's always a little bit of variation

14
00:00:39,010 --> 00:00:40,810
to application of the standard

15
00:00:40,810 --> 00:00:43,570
because that's what comes
with a human visual system.

16
00:00:43,570 --> 00:00:45,310
How do you reduce that variation

17
00:00:45,310 --> 00:00:47,130
that comes into a grading process,

18
00:00:47,130 --> 00:00:50,840
and accurately and consistently
grade the same diamond?

19
00:00:50,840 --> 00:00:52,700
I reached out to IBM Research

20
00:00:52,700 --> 00:00:54,849
and then taking our knowledge,

21
00:00:54,849 --> 00:00:56,250
inquired gemologists,

22
00:00:56,250 --> 00:00:58,955
who's been with GIA plus 40 plus years.

23
00:00:58,955 --> 00:01:01,400
We trained AI with it.

24
00:01:01,400 --> 00:01:04,620
The output went into the IBM Cloud.

25
00:01:04,620 --> 00:01:07,780
Our graders are our
best asset that we have.

26
00:01:07,780 --> 00:01:10,040
We will be able to
leverage them into areas

27
00:01:10,040 --> 00:01:12,380
that we haven't been
able to do that today.

28
00:01:12,380 --> 00:01:14,564
People are not afraid
about AI taking over jobs.

29
00:01:14,564 --> 00:01:16,070
In contrary to that,

30
00:01:16,070 --> 00:01:19,210
our teams are excited
about using the technology

31
00:01:19,210 --> 00:01:21,430
along the side of what they do today.

32
00:01:21,430 --> 00:01:23,050
Security is the heart of everything.

33
00:01:23,050 --> 00:01:24,980
Part of our requirements with IBM

34
00:01:24,980 --> 00:01:27,710
was to create this robust cloud solution

35
00:01:27,710 --> 00:01:30,304
that provides us a very secure,

36
00:01:30,304 --> 00:01:32,770
end-to-end encryption of data,

37
00:01:32,770 --> 00:01:33,820
that leaves GIA,

38
00:01:33,820 --> 00:01:35,000
that goes into the cloud,

39
00:01:35,000 --> 00:01:36,940
and it gets delivered back to it.

40
00:01:36,940 --> 00:01:38,760
Working with the ESA team has been great.

41
00:01:38,760 --> 00:01:42,270
They took time to
understand our requirement.

42
00:01:42,270 --> 00:01:44,300
How we are going to deploy the solutions

43
00:01:44,300 --> 00:01:46,680
around the world in different labs?

44
00:01:46,680 --> 00:01:49,560
They proposed architected solution

45
00:01:49,560 --> 00:01:52,600
that included Kubernetes clusters,

46
00:01:52,600 --> 00:01:54,290
as well as, different GPU's

47
00:01:54,290 --> 00:01:56,690
that allows us to throttle up or down

48
00:01:56,690 --> 00:02:00,490
as our volume of the data
process through the solution.

49
00:02:00,490 --> 00:02:02,120
It's been great working with that team

50
00:02:02,120 --> 00:02:05,200
and they bring a great deep
expertise in this area.

51
00:02:05,200 --> 00:02:08,760
This partnership has been
tremendous for GIA as well as IBM.

52
00:02:08,760 --> 00:02:12,730
The two organizations brings
a deep expertise in the areas

53
00:02:12,730 --> 00:02:16,766
that nobody else in our
industry could do it together.

54
00:02:16,766 --> 00:02:19,349
(gentle music)