1
00:00:02,480 --> 00:00:06,840
Speaker 1: Bloomberg Audio Studios, Podcasts, radio news.

2
00:00:07,320 --> 00:00:11,200
Speaker 2: This is Bloomberg Business Week with Carol Messer and tim

3
00:00:11,240 --> 00:00:15,520
Speaker 2: Stenoveek on Bloomberg Radio Sounds of Bono the Edge and

4
00:00:15,560 --> 00:00:17,960
Speaker 2: other members of You two, you know, wrapping up their

5
00:00:17,960 --> 00:00:20,680
Speaker 2: residency at the Stay one hundred and sixty six foot

6
00:00:20,720 --> 00:00:23,280
Speaker 2: tall two point three billion dollar orb in Las Vegas,

7
00:00:23,360 --> 00:00:24,239
Speaker 2: known as the Sphere.

8
00:00:24,280 --> 00:00:26,400
Speaker 3: Carol, you and I talk about it all.

9
00:00:26,320 --> 00:00:28,440
Speaker 1: The time, Rewhind because we want to be there.

10
00:00:28,200 --> 00:00:30,800
Speaker 2: Catching the show. Needless say, we haven't made it yet.

11
00:00:30,840 --> 00:00:32,960
Speaker 2: It's not looking so good, not yet, not yet.

12
00:00:33,080 --> 00:00:33,800
Speaker 1: You're still young.

13
00:00:34,159 --> 00:00:36,600
Speaker 2: Yeah, Well they're going to be done in a few days.

14
00:00:36,640 --> 00:00:38,720
Speaker 1: Well, there'll be somebody else there, that's true, all right.

15
00:00:38,760 --> 00:00:41,559
Speaker 1: Well maybe the next best thing I guess to not

16
00:00:41,840 --> 00:00:44,000
Speaker 1: getting there is getting some time with the company that's

17
00:00:44,000 --> 00:00:46,120
Speaker 1: the official tech partner for you two, helping to enable

18
00:00:46,159 --> 00:00:49,360
Speaker 1: the high def imagery. We're welcoming in our studio Jonathan Martin.

19
00:00:49,560 --> 00:00:52,159
Speaker 1: He's the president of the data management provider WEKCA and

20
00:00:52,200 --> 00:00:53,519
Speaker 1: as we said, he's here in studio.

21
00:00:53,560 --> 00:00:55,440
Speaker 4: Welcome, Welcome, thank you, great to be here.

22
00:00:55,440 --> 00:00:58,040
Speaker 1: Tell us about your company WEAKCA so WECA.

23
00:00:57,760 --> 00:01:03,000
Speaker 4: Is an AI native data platform that allows large AI

24
00:01:03,920 --> 00:01:05,160
Speaker 4: environments to be built.

25
00:01:05,480 --> 00:01:07,680
Speaker 1: So it was a year and a half ago would

26
00:01:07,680 --> 00:01:08,880
Speaker 1: you be talking about AI so much.

27
00:01:09,160 --> 00:01:11,800
Speaker 4: So we've been talking about AI for probably the last

28
00:01:11,840 --> 00:01:14,360
Speaker 4: five years, but it does seem to be reasonably hit

29
00:01:14,440 --> 00:01:17,240
Speaker 4: these days. So about two hundred and seventy five of

30
00:01:17,240 --> 00:01:20,040
Speaker 4: the world's largest companies are using Wekker today, eleven out

31
00:01:20,040 --> 00:01:23,080
Speaker 4: of the Fortune fifty, and one of those obviously was

32
00:01:23,520 --> 00:01:24,040
Speaker 4: the sphere.

33
00:01:24,120 --> 00:01:25,640
Speaker 1: What are companies mostly using you for?

34
00:01:26,040 --> 00:01:28,880
Speaker 4: So they're using is for building very large scale data pipelines.

35
00:01:28,920 --> 00:01:32,319
Speaker 4: So if you imagine the companies building these AI environments

36
00:01:32,600 --> 00:01:36,760
Speaker 4: are deploying very large volumes of GPUs, thousands, tens of

37
00:01:36,800 --> 00:01:39,520
Speaker 4: thousands of GPUs, and they want to be able to

38
00:01:39,560 --> 00:01:42,640
Speaker 4: serve data very very quickly, very very scalably, very very

39
00:01:42,640 --> 00:01:45,120
Speaker 4: efficiently to those GPUs to make sure that the GPUs

40
00:01:45,240 --> 00:01:47,400
Speaker 4: are running as fast as they can. So typically when

41
00:01:47,440 --> 00:01:50,760
Speaker 4: they deploy Wekka, they'll see that things like training times

42
00:01:50,760 --> 00:01:53,960
Speaker 4: for AI on the models will shrink, you know, anywhere

43
00:01:53,960 --> 00:01:56,040
Speaker 4: from ten to one hundred times. If you can imagine

44
00:01:56,160 --> 00:01:59,080
Speaker 4: what you could do with one hundred times more in

45
00:01:59,120 --> 00:02:01,280
Speaker 4: a day, it's a pretty incredible impact.

46
00:02:01,840 --> 00:02:05,880
Speaker 2: Okay, So privately held company privately company YEP raised a

47
00:02:05,920 --> 00:02:07,640
Speaker 2: lot of money last year, raised one hundred and thirty

48
00:02:07,680 --> 00:02:12,120
Speaker 2: five million dollars in a series defunding round what's the

49
00:02:12,160 --> 00:02:16,080
Speaker 2: plan for going public or an exit here for these investors.

50
00:02:16,160 --> 00:02:18,920
Speaker 4: So we're really focused on building the next great data

51
00:02:18,960 --> 00:02:21,040
Speaker 4: company to come out of Silicon Valley. I think we

52
00:02:21,080 --> 00:02:24,640
Speaker 4: have an absolutely incredible opportunity ahead of us. As you said,

53
00:02:25,040 --> 00:02:28,320
Speaker 4: AI is huge right now, but there are many, many

54
00:02:28,360 --> 00:02:30,400
Speaker 4: other sectors You've been talking about some of them earlier

55
00:02:30,440 --> 00:02:34,320
Speaker 4: this afternoon, like media and entertainment, life sciences, financial services

56
00:02:34,360 --> 00:02:36,000
Speaker 4: that are reimagining themselves.

57
00:02:36,040 --> 00:02:37,600
Speaker 3: How do you see AI playing a role in what

58
00:02:37,600 --> 00:02:38,080
Speaker 3: they're doing.

59
00:02:38,680 --> 00:02:42,600
Speaker 4: So, media and entertainment is leveraging AI massively. A lot

60
00:02:42,600 --> 00:02:48,320
Speaker 4: of generative is being used for character development, backgrounding, sequencing,

61
00:02:48,480 --> 00:02:50,720
Speaker 4: in between ing. So lots and lots of generative in

62
00:02:51,120 --> 00:02:52,320
Speaker 4: media and entertainment.

63
00:02:52,560 --> 00:02:54,600
Speaker 1: So I get it with you too, right. You helped

64
00:02:54,639 --> 00:02:57,280
Speaker 1: create the visuals around.

65
00:02:56,919 --> 00:02:59,799
Speaker 4: It, Yeah, so we completely did it. So two things.

66
00:02:59,800 --> 00:03:02,200
Speaker 4: So we gave them the platform on which they could

67
00:03:02,200 --> 00:03:04,880
Speaker 4: build those visuals. So you can imagine that that is

68
00:03:05,120 --> 00:03:08,280
Speaker 4: a very very unique environment for sixteen K screens one

69
00:03:08,360 --> 00:03:12,000
Speaker 4: hundred and sixty four thousand independent channels of audio. That's

70
00:03:12,160 --> 00:03:14,520
Speaker 4: that's targeted three seats at a time. So the sound

71
00:03:14,520 --> 00:03:18,239
Speaker 4: in there is absolutely beautiful, and they're streaming data about

72
00:03:18,240 --> 00:03:19,840
Speaker 4: four hundred and two gigabytes a seconds.

73
00:03:19,840 --> 00:03:21,640
Speaker 1: Tell us about some other customers, like, I get that,

74
00:03:21,680 --> 00:03:23,440
Speaker 1: but there's not a million spears.

75
00:03:23,160 --> 00:03:24,080
Speaker 4: There's not a million space.

76
00:03:26,040 --> 00:03:27,960
Speaker 1: But give us an idea of the kind of your

77
00:03:28,000 --> 00:03:31,080
Speaker 1: typical customer and what they're doing with your we're doing

78
00:03:31,080 --> 00:03:34,240
Speaker 1: with you with wekka, and what they're using what you provide.

79
00:03:33,800 --> 00:03:36,520
Speaker 4: Typically so many many other media and entertainment companies. So

80
00:03:36,520 --> 00:03:38,960
Speaker 4: a lot of the shows that Union families watch. A

81
00:03:39,000 --> 00:03:42,360
Speaker 4: lot of the broadcast studios in the sky are built

82
00:03:42,360 --> 00:03:46,160
Speaker 4: on Wekka. But we're also very strong in things like financial.

83
00:03:45,680 --> 00:03:47,400
Speaker 1: Services broadcast studios in the sky.

84
00:03:47,560 --> 00:03:49,960
Speaker 4: So when they're building, so for example, if you want

85
00:03:49,960 --> 00:03:52,960
Speaker 4: to go and build, if you've got you know, four

86
00:03:53,040 --> 00:03:55,320
Speaker 4: or five studios around the world, they pull all of

87
00:03:55,360 --> 00:03:58,000
Speaker 4: their content into into a cloud based studio. You have

88
00:03:58,120 --> 00:04:02,120
Speaker 4: maybe another thousand or two thousand reporters out there with

89
00:04:02,400 --> 00:04:05,200
Speaker 4: camera phones putting them all into the media base. So

90
00:04:05,240 --> 00:04:07,960
Speaker 4: a lot of broadcast networks are building these studios in

91
00:04:07,960 --> 00:04:11,000
Speaker 4: the sky where they're having people pull all the data

92
00:04:11,240 --> 00:04:14,200
Speaker 4: into into a cloud built on Wekka, and then they're

93
00:04:14,240 --> 00:04:18,080
Speaker 4: having the producers and everybody pull from those streams. Not

94
00:04:18,440 --> 00:04:21,640
Speaker 4: one like people people who you will definitely know.

95
00:04:21,880 --> 00:04:24,040
Speaker 3: Yes, Hey, I want to talk more about investors.

96
00:04:24,440 --> 00:04:30,080
Speaker 1: Use people like traditional networks, traditional news networks are not yet.

97
00:04:31,680 --> 00:04:31,920
Speaker 3: Checking.

98
00:04:32,000 --> 00:04:33,440
Speaker 4: All right, Well, you're in the building, so maybe it

99
00:04:33,520 --> 00:04:35,360
Speaker 4: exactly introduce me to your friends.

100
00:04:36,600 --> 00:04:39,799
Speaker 2: So the other other investors include Hewlett Packard Enterprise, HPE,

101
00:04:40,480 --> 00:04:45,000
Speaker 2: Micron Ventures. You got Nvidio in here, you got Qualcom Ventures, Samsung.

102
00:04:46,360 --> 00:04:48,800
Speaker 2: Are you using any of their hardware?

103
00:04:49,680 --> 00:04:51,880
Speaker 4: So we're software company, I know, but are you in

104
00:04:51,960 --> 00:04:54,800
Speaker 4: terms of like developing this stuff? So, but we run

105
00:04:54,960 --> 00:04:57,720
Speaker 4: as what we call reference architectures. So you can buy

106
00:04:57,800 --> 00:05:00,360
Speaker 4: Wekka from HP, you can buy it from the you

107
00:05:00,400 --> 00:05:01,760
Speaker 4: can buy it from Dell. You can buy it from

108
00:05:01,800 --> 00:05:02,359
Speaker 4: super Micro.

109
00:05:02,760 --> 00:05:03,920
Speaker 3: What is it? What do you mean by that? Is

110
00:05:03,960 --> 00:05:04,800
Speaker 3: it like white label?

111
00:05:05,080 --> 00:05:07,920
Speaker 4: So it's sold as a worker product and it's packaged

112
00:05:07,960 --> 00:05:10,840
Speaker 4: with hardware from Dell or hpoper Micro. You can also

113
00:05:10,880 --> 00:05:12,440
Speaker 4: buy the same product. And this is one of the

114
00:05:12,480 --> 00:05:15,160
Speaker 4: things that's extremely unique about its absolutely the same product

115
00:05:15,240 --> 00:05:17,800
Speaker 4: on any of your cloud marketplaces. So you can buy

116
00:05:17,839 --> 00:05:21,280
Speaker 4: it on AWS or Google or Azure or OCI.

117
00:05:21,600 --> 00:05:23,640
Speaker 3: Who's the biggest competitor you're concerned about.

118
00:05:24,320 --> 00:05:27,279
Speaker 4: M honestly ignorance at the moment, Like a lot of

119
00:05:27,320 --> 00:05:31,200
Speaker 4: people are just doing the same old, same old, same old.

120
00:05:31,200 --> 00:05:34,839
Speaker 4: These AI workloads are very, very different. They require extreme performance,

121
00:05:34,880 --> 00:05:38,320
Speaker 4: they require extreme scale, they require simplicity and shareability, and

122
00:05:38,360 --> 00:05:40,600
Speaker 4: they require you to have the ability to build these

123
00:05:40,720 --> 00:05:44,280
Speaker 4: pipelines across data centers and cloud environments.

124
00:05:43,800 --> 00:05:45,800
Speaker 1: A lot of extremes. And we've talked about this a lot,

125
00:05:45,839 --> 00:05:47,800
Speaker 1: whether it's autonomous vehicles and so on and so forth,

126
00:05:47,839 --> 00:05:51,599
Speaker 1: the use of energy and power to do all of this.

127
00:05:52,320 --> 00:05:55,279
Speaker 1: How are you guys focusing on this and making we

128
00:05:55,360 --> 00:05:56,599
Speaker 1: do it in a greener way?

129
00:05:56,839 --> 00:05:59,279
Speaker 4: So Sam Altman got on stage at Davos this year

130
00:05:59,320 --> 00:06:02,320
Speaker 4: and said that the world doesn't appreciate the power requirements

131
00:06:02,320 --> 00:06:05,520
Speaker 4: of AI and the AI revolution. That's absolutely what we see.

132
00:06:05,600 --> 00:06:07,920
Speaker 4: So these GPUs, if you have a thousand GPUs, they

133
00:06:07,920 --> 00:06:11,080
Speaker 4: consume about a megawatt of power. People are buying GPUs

134
00:06:11,120 --> 00:06:14,800
Speaker 4: these times in tens, maybe even hundreds of thousands, so massive, massive,

135
00:06:14,800 --> 00:06:18,800
Speaker 4: massive power requirements. So there's a big focus on sustainability

136
00:06:19,160 --> 00:06:22,120
Speaker 4: and how do we do this in a greener way. Interestingly,

137
00:06:22,360 --> 00:06:25,520
Speaker 4: our the series D that you mentioned was led by

138
00:06:25,680 --> 00:06:28,599
Speaker 4: Al Gore's fund, which is the Generation Investment Management Fund.

139
00:06:28,800 --> 00:06:30,720
Speaker 4: We were the second investment out of their Green Data

140
00:06:30,760 --> 00:06:35,080
Speaker 4: Fund because we help organizations massively reduce the amount of

141
00:06:35,080 --> 00:06:38,480
Speaker 4: infrastructury they require to get the same result. So typically

142
00:06:38,640 --> 00:06:40,960
Speaker 4: for every petabyte of wekka that you buy, you save

143
00:06:41,000 --> 00:06:43,680
Speaker 4: about two hundred and sixty tons of carbon dioxide emissions.

144
00:06:43,800 --> 00:06:46,880
Speaker 4: When we're talking about environments that are hundreds of petabytes

145
00:06:46,920 --> 00:06:49,479
Speaker 4: breaking into exobytes. Tho's on massive savings.

146
00:06:49,720 --> 00:06:52,560
Speaker 1: So you talked about kind of news, you talked about entertainment.

147
00:06:52,720 --> 00:06:56,360
Speaker 1: Where else is WEKA being used? Give me an idea.

148
00:06:56,520 --> 00:07:00,920
Speaker 4: Life sciences, so personal medicine, A lot of the MR

149
00:07:00,920 --> 00:07:03,880
Speaker 4: and A vaccines were developed on wekka. So computational chemistry,

150
00:07:03,960 --> 00:07:07,880
Speaker 4: structural biology, all of the FDA approvals we done on us. Again,

151
00:07:08,279 --> 00:07:12,240
Speaker 4: because we can massively compress walclock time. What may may

152
00:07:12,280 --> 00:07:14,960
Speaker 4: have taken you twelve days, we can now do in

153
00:07:15,000 --> 00:07:15,920
Speaker 4: four hours.

154
00:07:16,400 --> 00:07:19,400
Speaker 2: We love numbers here at Bloomberg. You're a private company,

155
00:07:19,440 --> 00:07:21,280
Speaker 2: but I'm going to look for some numbers here. We

156
00:07:21,400 --> 00:07:23,920
Speaker 2: just heard from Salesforce CEO Mark Bennioff, who said he's

157
00:07:23,960 --> 00:07:28,160
Speaker 2: excited about the all the spend going into it this year.

158
00:07:28,400 --> 00:07:30,640
Speaker 2: How much have you seen spend go up for wekka?

159
00:07:31,640 --> 00:07:34,480
Speaker 4: So we are tripling each year and of triple for

160
00:07:34,480 --> 00:07:37,560
Speaker 4: the last three years. Top line revenue, top line revenue.

161
00:07:37,720 --> 00:07:39,120
Speaker 3: They are, are you profitable?

162
00:07:40,120 --> 00:07:43,280
Speaker 4: We're a private company. That's okay.

163
00:07:43,360 --> 00:07:45,920
Speaker 1: You could tell us even if you're private, yes or

164
00:07:45,960 --> 00:07:46,520
Speaker 1: no question.

165
00:07:47,480 --> 00:07:50,440
Speaker 4: So we're focused on growth right now. We're focused on growth.

166
00:07:51,200 --> 00:07:54,040
Speaker 1: Fascinating. I'm quite not quite sure way to go because

167
00:07:54,040 --> 00:07:55,640
Speaker 1: I feel like I have a million questions, But having

168
00:07:55,760 --> 00:07:59,680
Speaker 1: said that, we have a smart investment in audience. You're

169
00:08:00,760 --> 00:08:03,920
Speaker 1: at least for the moment. But what is it I

170
00:08:03,920 --> 00:08:07,640
Speaker 1: don't know whether it's your AI exposure or you're exposure

171
00:08:07,680 --> 00:08:09,760
Speaker 1: to different industries. What is it that you think they

172
00:08:09,840 --> 00:08:12,200
Speaker 1: can kind of take away from this conversation that we're

173
00:08:12,200 --> 00:08:13,600
Speaker 1: having with you that you think that they should know.

174
00:08:13,560 --> 00:08:15,600
Speaker 4: About Probably the simplest way to think it doesn't have

175
00:08:15,600 --> 00:08:15,840
Speaker 4: to be.

176
00:08:15,760 --> 00:08:17,920
Speaker 1: Company specific, but more broadly what you are seeing.

177
00:08:18,000 --> 00:08:18,200
Speaker 3: Yeah.

178
00:08:18,280 --> 00:08:21,080
Speaker 4: So, so, first of all, we are very very early

179
00:08:21,120 --> 00:08:24,080
Speaker 4: in the AI revolution. We are you know, twenty twenty

180
00:08:24,080 --> 00:08:28,360
Speaker 4: three really saw the early adopters in the market, grabbing training,

181
00:08:28,600 --> 00:08:31,040
Speaker 4: grabbing in the world at large, the world at largest,

182
00:08:31,080 --> 00:08:35,360
Speaker 4: it's very very very early early adoptors. Even geographically it's

183
00:08:35,440 --> 00:08:38,520
Speaker 4: very very spotty. West coast of the US. AI is

184
00:08:38,600 --> 00:08:41,760
Speaker 4: very very hot, East coast not so much. You go

185
00:08:41,840 --> 00:08:43,959
Speaker 4: to Dubai, you go to Saudi, you go to q

186
00:08:44,160 --> 00:08:48,480
Speaker 4: eight incredible investment, You go to Taiwan, you go to Singapore,

187
00:08:48,600 --> 00:08:52,400
Speaker 4: you go to Korea. Incredibly AI in AI and AI

188
00:08:52,760 --> 00:08:56,120
Speaker 4: and not just not just AI, in all the infrastructure

189
00:08:56,160 --> 00:08:59,520
Speaker 4: to build an AI industry. So you go to Saudi,

190
00:08:59,559 --> 00:09:03,560
Speaker 4: they're building dozens of universities to churn out AI graduates.

191
00:09:04,160 --> 00:09:09,000
Speaker 4: So it's government level investment in the infrastructure. There is

192
00:09:09,120 --> 00:09:11,000
Speaker 4: the market or the market is.

193
00:09:11,080 --> 00:09:12,120
Speaker 1: Just got about fifteen seconds.

194
00:09:12,240 --> 00:09:14,400
Speaker 4: Yeah, super super early. So the best way to think

195
00:09:14,400 --> 00:09:18,600
Speaker 4: about Weka really is in AI AI environments. Wekka is

196
00:09:18,640 --> 00:09:21,160
Speaker 4: to data what Nvidia is to compute.

197
00:09:21,559 --> 00:09:25,040
Speaker 1: Very interesting. Come back soon and keep us updated. Jonathan Martin,

198
00:09:25,080 --> 00:09:26,960
Speaker 1: he's preident of Weka joining us here in studio