WEBVTT

1
00:00:00.000 --> 00:00:06.480
Diagnosed with cancer. I'm Chris Karragio, NBC News Radio, NBC News on

2
00:00:06.679 --> 00:00:12.519
CACAA Lomolinda sponsored by Teamsters Local nineteen
thirty two, protecting the Future of working

3
00:00:12.519 --> 00:00:22.440
Families, Teamsters nineteen thirty two.
Dot org. The information economy has a

4
00:00:22.559 --> 00:00:27.960
rod. The world is teeming with
innovation as new business models reinvent every industry

5
00:00:28.000 --> 00:00:33.320
industry. Inside Analysis is your source
of information and insight about how to make

6
00:00:33.359 --> 00:00:38.119
the most of this exciting new era. Learn more at inside analysis dot com,

7
00:00:38.159 --> 00:00:42.759
insideanalysis dot com. And now here's
your host, through Eric Kavanaugh.

8
00:00:47.359 --> 00:00:50.479
All Right, ladies and gentlemen,
Hello and welcome back once again to the

9
00:00:50.520 --> 00:00:54.560
only coast to coast show all about
the information economy. It's time for Inside

10
00:00:54.600 --> 00:00:57.960
Analysis. You're truly Eric Kavanaugh here
with an old buddy of mine in the

11
00:00:57.960 --> 00:01:02.600
industry, Todd Mostak. As with
Heavy AI these days, they've changed names

12
00:01:02.640 --> 00:01:06.040
a couple of times to be called
Omniside. Now it's heavy dot Ai,

13
00:01:06.159 --> 00:01:10.519
which is cool stuff. And they've
got a new offering and folks, this

14
00:01:10.599 --> 00:01:14.159
is really some interesting stuff. So
it's a database. It's at the GPU

15
00:01:14.239 --> 00:01:18.480
Accelerated database. Of course, you
know about all the different database technologies.

16
00:01:18.480 --> 00:01:23.079
There's a real explosion of database technologies
really about twelve fifteen years ago is when

17
00:01:23.159 --> 00:01:26.280
it started. And I remember I
had a theory at the time I floated

18
00:01:26.319 --> 00:01:29.920
by some of my friends. I
said that I think a lot of this

19
00:01:30.000 --> 00:01:34.719
open source database stuff is they're really
trying to hack away at Larry Ellison's corner

20
00:01:34.719 --> 00:01:38.319
of the market. Basically, they
want to rest control from Oracle, and

21
00:01:38.719 --> 00:01:41.959
hey, let's be honest. You
know, Oracle has some heavy handed practices

22
00:01:41.959 --> 00:01:45.760
over the years with lawyers and kind
of messing with folks, and it became

23
00:01:45.840 --> 00:01:49.359
painful and it caused a lot of
bad customer experience. I would say,

24
00:01:49.359 --> 00:01:53.439
not that the tech didn't work,
just that it was difficult to dealing with

25
00:01:53.439 --> 00:01:56.879
the licensure and all this stuff.
And now you have this whole I mean

26
00:01:57.319 --> 00:02:00.560
explosion of database types, all these
open source thingsdatabases, and one of them

27
00:02:00.799 --> 00:02:06.159
is heavy Heavy AI. It's a
database and heavy IQ is a new offering.

28
00:02:06.400 --> 00:02:07.560
So we're going to talk to Todd
about what this stuff is and what

29
00:02:07.599 --> 00:02:12.000
it means and how you could use
it. Frankly, so Tom want you

30
00:02:12.000 --> 00:02:15.879
to start us off with what is
the platform itself, Heavy AI, and

31
00:02:15.919 --> 00:02:20.919
what are you announcing now soome Yeah, well, first off, thanks for

32
00:02:20.960 --> 00:02:23.960
having me Eric, It's great to
get back. So yeah. Yeah,

33
00:02:24.039 --> 00:02:28.960
So you know Heavy AI, so
Heavy AI as a platform, we are

34
00:02:28.960 --> 00:02:34.599
a big data analytics platform. Our
kind of claim to fame is that we

35
00:02:34.639 --> 00:02:38.840
are GPU accelerated, So we leverage
the massive parallel processing power of GPUs,

36
00:02:38.960 --> 00:02:43.599
all those thousands of cores and terabytes
of memory done with a second of memory

37
00:02:43.599 --> 00:02:46.599
down with that you have on these
cards originally designed to render video games obviously

38
00:02:46.639 --> 00:02:52.520
now kind of part and parcel and
core to the machine learning and AI revolution.

39
00:02:53.240 --> 00:02:57.000
So we use all that supercomputing bandwidth
to good effect to actually process the

40
00:02:55.960 --> 00:03:00.560
EAM much faster. In particular,
run SEQL queries of billions of records in

41
00:03:00.599 --> 00:03:06.400
milliseconds without needing to downsample, to
pre index, to pre aggregate your data.

42
00:03:06.520 --> 00:03:09.639
We can use kind of the GPU's
power to do smart roof force across

43
00:03:09.680 --> 00:03:14.639
massive data sets and allowing you to
get kind of interactive real time answers,

44
00:03:15.400 --> 00:03:19.039
you know, not waiting for tens
of seconds, minutes, hours, depending

45
00:03:19.039 --> 00:03:22.680
on your database. You get questions, you get answers back immediately. We

46
00:03:22.719 --> 00:03:25.280
can also visualize those SEQL results in
line. We have our own front end

47
00:03:25.319 --> 00:03:30.319
called heavy immerse. So heavy dB
is the database engine running on GPUs because

48
00:03:30.599 --> 00:03:35.120
one of the killer apps of having
that kind of low latency and that kind

49
00:03:35.120 --> 00:03:39.599
of speed interactivity was enabling interactive visualization. So imagine real time data, you

50
00:03:39.639 --> 00:03:44.919
know, millions or billions of call
records, social media data, you know,

51
00:03:44.960 --> 00:03:49.840
whether it's Twitter now x or anything
else, whether it's you know,

52
00:03:49.879 --> 00:03:54.639
oil and gas, whether it's it's
cybersecurity, you know, logs on a

53
00:03:54.680 --> 00:03:59.400
server. We can process all that
data, visualize it. We can actually

54
00:03:59.479 --> 00:04:01.319
use the gpu is to render that
data. So if you want to display

55
00:04:01.319 --> 00:04:04.520
that data, g you spatially put
billions of points on the map. And

56
00:04:04.599 --> 00:04:08.479
so a lot of our customers say, you know, this is the core

57
00:04:08.520 --> 00:04:13.240
of it is fast database front end
is very interactive. Some people have called

58
00:04:13.240 --> 00:04:16.720
it Tableau on steroids, but we
have people using us both as a database

59
00:04:16.800 --> 00:04:21.600
and kind of a full stack visual
analytics platform. Well it's interesting because you

60
00:04:23.240 --> 00:04:29.560
attacked what is a fairly well populated
space already but with something new, and

61
00:04:29.600 --> 00:04:31.600
you look at like what a cloud
Darra and a Hortant works were doing,

62
00:04:31.839 --> 00:04:35.319
and initially, of course that was
all hdfs, which is a filesystem it's

63
00:04:35.319 --> 00:04:39.120
not a database, it's a filesystem, which is why they ran into so

64
00:04:39.160 --> 00:04:42.600
many problems. I think that was
one of the the achilles heels, if

65
00:04:42.639 --> 00:04:45.879
you will, of that whole approach, and that whole movement was that HTFS

66
00:04:46.000 --> 00:04:51.079
is a filesystem. Many of the
value propositions for that approach kind of dissipated,

67
00:04:51.240 --> 00:04:56.720
especially when cloud storage got so cheap, so there were some flaws in

68
00:04:56.759 --> 00:04:59.240
the architecture that. You know,
they've done a really good job to pivot,

69
00:04:59.480 --> 00:05:00.879
so they're doing much more interesting stuff
now. Then you have all the

70
00:05:01.000 --> 00:05:05.120
historical players, like your terror data
is your verticas then of course Oracle and

71
00:05:05.160 --> 00:05:11.839
IBM and all these guys. But
those relational database structures are good in certain

72
00:05:12.040 --> 00:05:15.040
use cases, but they're not good
in certain others. And a couple domains

73
00:05:15.079 --> 00:05:19.800
in particular cause them real trouble.
One is time series and one is geospatial.

74
00:05:20.240 --> 00:05:24.759
And if you bring in both of
those things, you know, architecture

75
00:05:24.839 --> 00:05:29.319
is really important for being able to
process that information, for being able to

76
00:05:29.399 --> 00:05:32.480
parallelize it, as you suggest,
being able to get some meaningful insight out

77
00:05:32.519 --> 00:05:36.000
of it. And I think that's
where your sweet spot is right is because

78
00:05:36.040 --> 00:05:43.240
you have this GPU foundation because you
can parallelize a lot of this heavy lifting,

79
00:05:43.639 --> 00:05:46.519
if you will. Maybe that's where
the name came from for heavy AI.

80
00:05:46.279 --> 00:05:50.519
You're able to tackle some of these
really dense challenges, you know,

81
00:05:50.560 --> 00:05:54.959
think life sciences. Like I say, I think the nexus of geospatial and

82
00:05:55.040 --> 00:05:59.800
time series, you're able to do
that much better than some traditional relational database

83
00:05:59.800 --> 00:06:02.480
could do, right, correct,
you know, absolutely, you know,

84
00:06:02.560 --> 00:06:08.759
I think relational databases tend to be
grab bags of functionality, right, and

85
00:06:08.920 --> 00:06:11.839
and that's you know, SEQL is
a very powerful versual language. We also

86
00:06:11.959 --> 00:06:15.759
speak SQL. However, we've really
focused on those use cases that we think

87
00:06:15.920 --> 00:06:21.160
need kind of that real time performance
uh temporal and geospatial or space show temporal

88
00:06:21.240 --> 00:06:26.160
use cases kind of first and foremost
among them, because like you know,

89
00:06:26.199 --> 00:06:30.839
we see that our customers and often
you know, they're not working in the

90
00:06:30.879 --> 00:06:34.639
world of like old school tph Let
me join my line line item to my

91
00:06:34.800 --> 00:06:38.759
orders table, to my customer's table, right, this is kind of real

92
00:06:38.759 --> 00:06:43.160
time logs they're looking at. You
know. For example, one use case

93
00:06:43.199 --> 00:06:47.600
we have with a customer illegal fishing, right like looking at on the federal

94
00:06:47.639 --> 00:06:51.160
side, looking at where you know, ships may be crossing and areas where

95
00:06:51.160 --> 00:06:55.920
they shouldn't be and actually doing real
time analysis and figuring out, okay,

96
00:06:55.920 --> 00:07:00.000
these ships turned off their beacons,
and so you're not only doing temporal now

97
00:07:00.240 --> 00:07:02.639
where these ships are going, or
doing spatial analysis. And to be able

98
00:07:02.680 --> 00:07:08.279
to run that over billions of records
with like very fast, high velocity incoming

99
00:07:08.319 --> 00:07:12.879
feeds and looking at historical data,
that's just something that even you know,

100
00:07:12.959 --> 00:07:16.600
even a performance a redshift or a
big barrier or snowflake, these databases just

101
00:07:16.639 --> 00:07:21.240
don't scale to. And so I
think we've found kind of our core differentiator,

102
00:07:21.519 --> 00:07:25.959
and it really is real time data
often with space shift, temporal and

103
00:07:26.120 --> 00:07:30.959
just today's databases aren't really built to
handle those kind of workloads and scales right

104
00:07:30.000 --> 00:07:34.040
well, you know your snowflake.
So the world certainly are designed for structured

105
00:07:34.120 --> 00:07:39.639
data, which I've always believed really
means data in relational tables essentially, that's

106
00:07:39.639 --> 00:07:43.199
what they mean by the term structured. And you know, data breaks is

107
00:07:43.199 --> 00:07:46.600
along the same lines, maybe a
slightly different approach to getting there, but

108
00:07:46.639 --> 00:07:53.319
there's still similar engines. And what
you've done is special because again you've you've

109
00:07:53.399 --> 00:07:59.519
targeted some of these much more complex
use cases that would require tremendous compute power

110
00:07:59.560 --> 00:08:05.199
fromditional CPUs, so much so that
it's just not really economically feasible, and

111
00:08:05.240 --> 00:08:09.319
it's not even really compute intensive feasible
in a sense. Is that about riting?

112
00:08:09.959 --> 00:08:13.720
Yeah, that's that's absolutely correct.
And then the thing about CPUs,

113
00:08:13.800 --> 00:08:16.839
right, it's also about compute density
in the sense of yes, you could

114
00:08:16.959 --> 00:08:20.720
pound four pound scale up a CPU
cluster to have equal compute to a GPU

115
00:08:20.759 --> 00:08:24.040
cluster. By that time you're at
hundreds of nodes, right, And then

116
00:08:24.079 --> 00:08:26.800
you got to think about interconnect You
got to think about how you're sharding your

117
00:08:26.879 --> 00:08:33.120
data. There's so many inefficiencies of
going to a massive scale out cluster versus

118
00:08:33.159 --> 00:08:35.679
scaling up. And we can also
get a multi noode, but first and

119
00:08:35.720 --> 00:08:39.759
foremost we're scaling up within even a
single machine multigpus each with thousands of course,

120
00:08:41.080 --> 00:08:45.120
that gives us a lot of processing
capability, and the bandwidth between these

121
00:08:45.159 --> 00:08:48.840
GPUs, if you're familiar with some
of the nvidia's mv link capability can be

122
00:08:48.919 --> 00:08:52.600
measured in you know, many hundreds
of gigabytes a second, going into the

123
00:08:52.639 --> 00:08:56.960
terabytes a second range, and so
that's a whole different architecture, right,

124
00:08:56.000 --> 00:09:00.960
where you can actually move data around
very quickly Versus your two hundred nodes of

125
00:09:00.960 --> 00:09:05.080
impula on some commodity cluster. That
gets very difficult to actually get that kind

126
00:09:05.080 --> 00:09:09.600
of performance just because you're spending most
of your time shipping data between nodes and

127
00:09:09.679 --> 00:09:13.480
not actually doing kind of the core
analytics operations. Yeah, that's pretty interesting.

128
00:09:13.559 --> 00:09:18.360
So the n v M e SSDs
is that part of your vision?

129
00:09:18.399 --> 00:09:20.639
Is that part of what makes the
magic happen here? Yeah, we're getting

130
00:09:20.639 --> 00:09:24.279
more into that world, right,
So you know, we are I call

131
00:09:24.279 --> 00:09:28.360
it in memory light in the sense
that we do persist a disc. You

132
00:09:28.399 --> 00:09:31.360
can we can page in and out
of disc. Obviously, our systems happiest

133
00:09:31.360 --> 00:09:35.919
when we can at least cash in
CPU memory kind of the core data that

134
00:09:35.919 --> 00:09:41.799
we're actively analyzing. But we're actually
working with some partners on the storage side,

135
00:09:41.120 --> 00:09:45.960
and there's some innovations called GP directs
and GP file system from the Nvidia

136
00:09:46.000 --> 00:09:50.799
side where we can read straight from
the MVMME device and without with bypassing effectively

137
00:09:50.840 --> 00:09:58.399
bypassing the CPU and going straight to
mv ME over the PCI PCI bus,

138
00:09:58.399 --> 00:10:01.799
and you can get some tremendous performance
and in fact you can even have remote

139
00:10:01.799 --> 00:10:05.639
storage these days with you know,
these flash blades and be pulling at hundreds

140
00:10:05.639 --> 00:10:07.960
of gigabytes a second, which is
kind of unimaginable, you know, from

141
00:10:09.039 --> 00:10:11.879
storage, and so all of a
sudden, your storage can start looking at

142
00:10:11.879 --> 00:10:15.559
memory, except you have potentially petabytes
on tap. Yeah. Well, the

143
00:10:15.559 --> 00:10:18.559
reason I'm bringing this up is because
we have another company we're working with,

144
00:10:18.639 --> 00:10:20.600
Ocient. I don't know if you've
come up, Yeah, I notice it,

145
00:10:22.039 --> 00:10:26.440
But they really thought through what these
NVM solid state DROGECT and in particularly

146
00:10:26.480 --> 00:10:28.919
the interface, which I think is
what you're kind of speaking to now,

147
00:10:28.960 --> 00:10:33.639
the ability to just pull in massive
amounts of data and what they call it

148
00:10:33.759 --> 00:10:37.679
is a is a compute adjacent storage
is kind of how they're referring to it.

149
00:10:37.679 --> 00:10:41.840
It sounds very similar to what you're
talking about, and it is.

150
00:10:41.240 --> 00:10:48.240
It's orders of magnitude greater in terms
of possible throughput. So it just changes

151
00:10:48.279 --> 00:10:50.639
the entire game in terms of how
you think about what you're going to do

152
00:10:50.679 --> 00:10:54.000
with this data and how you think
about setting up the architecture. And that's

153
00:10:54.000 --> 00:10:58.960
where I see you guys doing something
very similar in that you know you have

154
00:10:58.639 --> 00:11:05.480
specif typically addressed these heavy duty use
cases, which is also what Ocean is

155
00:11:05.519 --> 00:11:11.360
doing, and you're conquering it really
with the GPU accelerator, but also just

156
00:11:11.559 --> 00:11:16.000
a different architectural approach to solving the
problem, right, Yeah, exactly,

157
00:11:16.039 --> 00:11:20.440
And I think you know, these
new problems and new hardware architectures really do

158
00:11:20.919 --> 00:11:26.240
the new thinking, right. Like
you know, most common kind of data

159
00:11:26.279 --> 00:11:31.279
warehouses were built in an age where
you know, interconnect was was very slow,

160
00:11:31.320 --> 00:11:33.559
where storage was very slow. You
know, all these things are in

161
00:11:33.720 --> 00:11:37.440
CPUs, we're kind of getting faster
at a very small fixed rate per year.

162
00:11:37.879 --> 00:11:43.000
All of that's been kind of up
ended with GPUs with fast you know,

163
00:11:43.200 --> 00:11:46.759
whether fast fabric connect right, whether
it's internal, external to machine mv

164
00:11:46.879 --> 00:11:50.960
link, PC I four, pc
I five, And so I think it

165
00:11:52.039 --> 00:11:56.240
demands different approaches and how you tackle
this with software and in terms of deployment

166
00:11:58.320 --> 00:12:01.919
on prem also in the c what
are the options for deployment. Yeah,

167
00:12:01.960 --> 00:12:05.240
So this is I think actually been
one of the blessings. You know,

168
00:12:05.080 --> 00:12:09.559
heavy AI we were mat D back
in the day. You know, we

169
00:12:09.639 --> 00:12:11.759
came into a world where there was
very little cloud CHAPU compute, and so

170
00:12:11.799 --> 00:12:16.600
that forces us to develop first and
foremost, foremost on prim options. Obviously,

171
00:12:16.639 --> 00:12:20.120
now almost all our customers run the
cloud, except we do have the

172
00:12:20.120 --> 00:12:24.320
ability to run fully hybrid, to
run on prem, to run cloud where

173
00:12:24.440 --> 00:12:28.000
Docker rise. We can also run
bare metal. We can run fully offline

174
00:12:28.000 --> 00:12:31.519
and air gaps, which is important
for some of our customers. And so

175
00:12:31.879 --> 00:12:37.000
we're pretty versatile in that respect versus
being pure SaaS Micro service Play, which

176
00:12:37.080 --> 00:12:41.440
kind of hms you in a little
bit in terms of your deployment options.

177
00:12:41.039 --> 00:12:46.080
So I'm one of these dorks who
sits around and thinks about this stuff all

178
00:12:46.120 --> 00:12:48.279
the time. And you know,
I've said for a long time now that

179
00:12:48.399 --> 00:12:52.519
on Prem's demise you know, has
been exaggerated, or the rumors of on

180
00:12:52.559 --> 00:12:56.360
prem's demise have been exaggerated. It's
going to be here for a while.

181
00:12:56.840 --> 00:13:00.879
I think you're going to see a
bit of a movement back for sensitive data

182
00:13:00.919 --> 00:13:03.759
for PII, but also for lms, you know, to be able to

183
00:13:05.039 --> 00:13:07.960
for big organizations to be able to
do that stuff very effectively. You know,

184
00:13:07.960 --> 00:13:11.080
they're going to probably want to have
a lot of that IP stuff in

185
00:13:11.240 --> 00:13:16.279
house and basically recast their on prem
data center to be able to do new

186
00:13:16.320 --> 00:13:20.519
and different interesting things. But I
just wonder about you know, the full

187
00:13:20.519 --> 00:13:24.320
cloud environment because as I think about
it, all, you know, we've

188
00:13:24.360 --> 00:13:28.440
gone from the monolith to the containers, which is a complete departure. It's

189
00:13:28.440 --> 00:13:33.279
a whole new way of doing things
right, and there are trade offs in

190
00:13:33.399 --> 00:13:39.639
going fully containerized. I joked that
we sacrificed what is it, we sacrifice

191
00:13:39.679 --> 00:13:41.960
state at the altar of scale with
some of this stuff. And yeah,

192
00:13:43.080 --> 00:13:46.639
you can scale out and then scale
back down and that's very useful, but

193
00:13:46.720 --> 00:13:50.320
you do have to to manage states
some other way, and it is a

194
00:13:50.360 --> 00:13:54.440
bit of a it's a bit of
an overkill in certain use cases, in

195
00:13:54.480 --> 00:13:58.000
certain situations. But what do you
think about all that? Yeah, you

196
00:13:58.039 --> 00:14:01.480
know, I think the you know, containeration I think is generally a good

197
00:14:01.519 --> 00:14:05.159
thing. Most of our customers do
run containerized. But I think the push

198
00:14:05.200 --> 00:14:11.159
into extreme microservices, I mean that
can make sense if you're running at Facebook

199
00:14:11.240 --> 00:14:13.399
or metascale, right, I mean
obviously there are a lot of advantage of

200
00:14:13.399 --> 00:14:18.440
that, but the amount of complexities, deployment, of development of often overhead

201
00:14:18.759 --> 00:14:22.519
just in terms of performance upgoing,
you know, with the micro services approach

202
00:14:22.519 --> 00:14:28.120
of all these services talking with each
other, redundantly calculating things, you know,

203
00:14:28.200 --> 00:14:31.000
we've we're not quite a monolith,
but by having our core databases one

204
00:14:31.120 --> 00:14:35.039
architecture very close to the metal,
we think that's been a big part of

205
00:14:35.080 --> 00:14:39.039
our kind of performance wins versus parceling
it up into you know, you know,

206
00:14:39.159 --> 00:14:43.360
tens or even hundreds of different micro
services. So I think there's always,

207
00:14:43.480 --> 00:14:45.759
you know, there's always a pendulum, just like there is with you

208
00:14:45.799 --> 00:14:48.960
know, cloud to on prem and
then back to hybrid. I think the

209
00:14:50.840 --> 00:14:54.159
market finds equilibrium at some point,
and obviously extremes, you know, often

210
00:14:54.200 --> 00:14:58.519
don't make sense. So I think
we're at a people are a little bit

211
00:14:58.519 --> 00:15:01.240
more realistic and also realistic about the
privacy piece of running on prem and also

212
00:15:01.679 --> 00:15:05.200
sometimes the cost advantages. We've had
customers starting the cloud, but then as

213
00:15:05.240 --> 00:15:09.320
they kind of deepen their investment and
you know, heavy and big data analytics

214
00:15:09.320 --> 00:15:13.759
in general, you know, they
want to run on prem just because the

215
00:15:13.840 --> 00:15:18.120
costs the economics makes sense, right, No, that makes complete sense,

216
00:15:18.159 --> 00:15:20.039
and you know, just real quick
before the first break comes up. Here,

217
00:15:20.080 --> 00:15:24.159
there's also this concept of the modern
data stack, which I saw spring

218
00:15:24.240 --> 00:15:28.600
up I guess a couple of years
ago. Now it kind of feeds into

219
00:15:28.639 --> 00:15:31.360
this that you know, again,
you have sort of best of breed components

220
00:15:31.360 --> 00:15:35.000
that you string together in order to
have a very robust solution. Well,

221
00:15:35.000 --> 00:15:37.039
one of my best friends in the
industry, he passed away last year,

222
00:15:37.080 --> 00:15:41.000
unfortunately, but a super smart guy, and I remember talking to him a

223
00:15:41.000 --> 00:15:43.240
couple of years ago. He's like, yeah, it's modern data stack.

224
00:15:43.840 --> 00:15:46.919
It tends to be strung together with
a couple of frail lines of code.

225
00:15:46.919 --> 00:15:50.039
And I'm not a big fan of
that because if any point along the way

226
00:15:50.120 --> 00:15:52.440
it collapses, you've got yourself a
big problem. Right, And it seems

227
00:15:52.480 --> 00:15:56.840
to me that you've got enough heavy
lifting again in the database and in your

228
00:15:56.960 --> 00:16:02.000
environment, and you can visualize in
the same environment with the same technology that

229
00:16:02.080 --> 00:16:06.519
you're kind of circumventing some of those
challenges that about, right, I would

230
00:16:06.519 --> 00:16:10.240
say, so. I mean,
I think that's one of the potential attractions

231
00:16:10.279 --> 00:16:14.799
for customers to our platform, but
also something that we have to educate the

232
00:16:14.840 --> 00:16:17.360
market on. You know, everybody's
thinking, okay, we need to have

233
00:16:17.960 --> 00:16:21.240
all this piece wise modern data stack, and we go in and say,

234
00:16:21.279 --> 00:16:23.799
hey, actually, you know,
we don't want our place necessarily your data

235
00:16:23.840 --> 00:16:27.120
warehouse a record or your data leak, but for your last mile analytics.

236
00:16:27.120 --> 00:16:30.799
We can actually do the sequel,
the visualization, even a lot of the

237
00:16:30.879 --> 00:16:34.720
data science work all in one stack. And you know, you're seeing the

238
00:16:34.799 --> 00:16:37.840
data through a single pane of glass, and we can pull from your data

239
00:16:37.840 --> 00:16:42.240
bricks, we can pull from your
Partka store. It's people say, it's

240
00:16:42.320 --> 00:16:47.360
much simpler to deploy and think about, and it's less headaches. Obviously if

241
00:16:47.360 --> 00:16:52.879
people are steeped in the notion that
you have to you know, parcel out

242
00:16:52.960 --> 00:16:56.080
your data stack into all these different
services. Obviously that can be a challenge

243
00:16:56.120 --> 00:16:57.639
selling. But we found that,
you know, customers get a lot of

244
00:16:57.679 --> 00:17:02.759
value out of having one kind of
a single platform to choke if you will.

245
00:17:02.799 --> 00:17:06.599
But you know, having all their
data in one place, with all

246
00:17:06.640 --> 00:17:11.960
the efficiencies both operationally and from a
performance and analytics perspective that you gain from

247
00:17:12.279 --> 00:17:18.039
having data consolidation versus many different silos. Yeah. Well, and I'm curious

248
00:17:18.079 --> 00:17:21.920
to know In the next segment,
we'll dive deeper into this. You know,

249
00:17:21.960 --> 00:17:25.720
how you populate this thing. How
much time does that take? You've

250
00:17:25.720 --> 00:17:30.880
talked about serving as a front end
with a data lake or a data warehouse

251
00:17:30.920 --> 00:17:33.559
in the background. That's also pretty
interesting because you know, I get that

252
00:17:34.039 --> 00:17:40.079
one environment for doing the processing.
If you could parallelize very well, that's

253
00:17:40.119 --> 00:17:44.119
a very compelling storyline because you're not
bouncing from system to system. You're in

254
00:17:44.200 --> 00:17:48.359
one environment and it's designed to handle
all these different queries that you do and

255
00:17:48.799 --> 00:17:52.599
give you the visualization all that stuff. So it's a very interesting, bold

256
00:17:52.680 --> 00:17:55.880
approach, I would say, but
don't touch that, tot. Folks will

257
00:17:55.880 --> 00:18:07.200
be right back. You're listening to
Inside Analysis. Welcome back to Inside Analysis.

258
00:18:07.720 --> 00:18:12.519
Here's your host, Eric Tabanaugh.
All right, folks, back here

259
00:18:12.519 --> 00:18:18.000
at Inside Analysis talking to Todd Mastak
with Heavy dot AI doing the heavy lifting

260
00:18:18.079 --> 00:18:22.160
for you, folks. I always
joke about the heavylifting and data warehousing and

261
00:18:22.839 --> 00:18:26.519
analytics and data science of course,
and one of the cool things about heavy

262
00:18:26.559 --> 00:18:32.920
AIS they talk about really bridging the
gap between those worlds because and I've always

263
00:18:32.920 --> 00:18:34.519
been fascinated by this. I'd like
to get your thoughts on this, Todd.

264
00:18:34.519 --> 00:18:40.480
But from what I've seen, data
science teams are very separate from data

265
00:18:40.480 --> 00:18:45.079
warehousing teams, and I have no
idea why that would remain the case as

266
00:18:45.160 --> 00:18:48.160
data science matures. I would think
you'd want these people talking to each other

267
00:18:48.200 --> 00:18:52.000
and working together because the same goals. Why would you have a completely discrete

268
00:18:52.000 --> 00:18:56.200
team for your data warehouse versus your
data science team. And it's just organizationally

269
00:18:56.200 --> 00:19:00.000
how things evolved. But I don't
think it makes a lot of sense.

270
00:19:00.359 --> 00:19:02.680
So maybe talk about that for a
minute and then we'll get into how you

271
00:19:02.680 --> 00:19:06.640
load this thing. Go ahead,
Yeah, it is an interesting phenomenon,

272
00:19:06.680 --> 00:19:08.880
and you can explain why it happened. But I do think in many cases,

273
00:19:08.880 --> 00:19:14.480
for many organizations, it's suboptimal to
have these you know, silot orgs,

274
00:19:14.519 --> 00:19:18.119
you know, data science versus you
know, your your database or your

275
00:19:18.200 --> 00:19:22.839
data engineering or your analytics teams,
because they're often working on roughly the same

276
00:19:22.880 --> 00:19:26.400
thing across purposes, and they distrust
each other, you know, and have

277
00:19:26.519 --> 00:19:30.799
different systems and different you know,
it can be kind of a mess.

278
00:19:30.799 --> 00:19:34.680
Some organizations handle it much better than
others. You know. We we're not

279
00:19:34.799 --> 00:19:40.880
claiming to be the panacea here,
but certainly we've had you know, one

280
00:19:40.880 --> 00:19:47.279
of our customers, a big hardware
manufacturer out in California, has seen themn's

281
00:19:47.319 --> 00:19:52.319
success with actually having both kind of
production operational teams hitting the visualizations, running

282
00:19:52.319 --> 00:19:56.839
BASER re COORSE and SQL, and
having data science teams working on the same

283
00:19:56.920 --> 00:20:00.680
data, cleaning the same data on
the same instance. Since it's you know,

284
00:20:00.680 --> 00:20:03.759
and they say it's a huge it's
a huge gain, and it actually

285
00:20:03.799 --> 00:20:08.000
these teams are working more closely together
than they ever have before. They're sharing

286
00:20:08.000 --> 00:20:12.680
insights and even though they're very different
personas, And some people are sitting there

287
00:20:12.839 --> 00:20:17.960
in a jupinter notebook which is attached
to heavy via our Python connector, and

288
00:20:18.440 --> 00:20:21.440
some people are using Immerse and then
other people might bull Tableau on top.

289
00:20:21.640 --> 00:20:25.200
So yeah, that's the dream.
It doesn't always happen that way, but

290
00:20:25.279 --> 00:20:29.240
you know, we like to think
that we're helping the two sides of the

291
00:20:29.279 --> 00:20:32.720
aisle get a little closer together.
Yeah. Well, and there's this jen

292
00:20:32.759 --> 00:20:37.640
Ai component too, So your new
offering, heavy IQ is allowing your customers

293
00:20:37.640 --> 00:20:42.319
to leverage the Geni's capabilities. And
we were talking before the show about RAG

294
00:20:42.400 --> 00:20:48.240
models, retrieval, logmented generation,
which I think are going to be a

295
00:20:48.359 --> 00:20:52.440
huge area of focus from here on
out. Basically, I think they're going

296
00:20:52.519 --> 00:20:57.799
to be the mechanism by which very
serious companies in the space do their data

297
00:20:57.880 --> 00:21:03.720
quality. Their data got even some
security is going to get baked in there,

298
00:21:03.440 --> 00:21:08.480
which is very different because in a
traditional data warehousing world, the data

299
00:21:08.519 --> 00:21:12.759
governance, the data quality that's baked
in in different ways, and you have

300
00:21:12.759 --> 00:21:17.720
different kinds of guardrails. But it's
like, if the LM is the interaction

301
00:21:17.839 --> 00:21:22.640
point for all of your questions,
well, now I can focus on building

302
00:21:22.640 --> 00:21:26.000
my guardrails inside the RAG model of
the LLM. But what do you think

303
00:21:26.039 --> 00:21:30.119
about that change? Because the other
thing I look at is the capacity of

304
00:21:30.160 --> 00:21:36.599
these engines, not just the text
generative capability, but their pattern recognition capabilities

305
00:21:36.640 --> 00:21:41.680
for example, and the mechanism by
which they choose what they choose is pretty

306
00:21:41.680 --> 00:21:45.000
interesting. It's not an exact science
yet. And you know a lot of

307
00:21:45.039 --> 00:21:49.319
people like talk to Joke, including
people who are very experienced in this industry.

308
00:21:49.559 --> 00:21:53.000
I really don't know exactly how it
works. It's like, well,

309
00:21:53.200 --> 00:21:56.200
you know, how's that going to
work? But what do you think about

310
00:21:56.240 --> 00:22:00.039
all of that? I mean,
are we really looking at upending many of

311
00:22:00.079 --> 00:22:03.119
the traditional ways we've done things over
the past twenty thirty years? What do

312
00:22:03.160 --> 00:22:07.599
you think? I think it will
upend things, and it's you know,

313
00:22:07.640 --> 00:22:11.599
it's going to be faster and slower
than anyone kind of imagines. I mean,

314
00:22:11.640 --> 00:22:15.839
I think faster in the sense of
like wow. You know, two

315
00:22:15.920 --> 00:22:18.640
years ago or a year and a
half ago, there was no shate GPT,

316
00:22:18.359 --> 00:22:22.279
you know, generative AI was being
talked about. That it was people

317
00:22:22.319 --> 00:22:25.680
who were just kind of putting their
feet in the water. And today everybody's

318
00:22:25.680 --> 00:22:29.160
thinking about everybody's using it, you
know. I think it's the whole high

319
00:22:29.200 --> 00:22:32.440
cycle thing. Right. There's a
lot of power you can get immediately,

320
00:22:32.559 --> 00:22:37.839
there's a lot of problems. You
know, the non determinism and the fuzziness

321
00:22:37.480 --> 00:22:41.039
is both a blessing and a curse
of it, right. You know,

322
00:22:41.119 --> 00:22:44.519
a lot of the work that we've
done in heavy IQ is how can we

323
00:22:44.599 --> 00:22:48.240
take these kind of you know,
this neveroless power that seems to have fallen

324
00:22:48.279 --> 00:22:52.440
from the sky. It's almost like
alien technology, and how can we like

325
00:22:52.519 --> 00:22:56.279
ballow it up into something that's predictable, that's reliable, that gives people the

326
00:22:56.319 --> 00:23:00.359
answers they want, that doesn't hallucinate
or almost never hallucinates. And how do

327
00:23:00.400 --> 00:23:03.279
we treat it like an engineering thing. We have unit tests around it,

328
00:23:03.839 --> 00:23:07.119
and what can we put in the
pipeline that's deterministic. I think sometimes people

329
00:23:07.119 --> 00:23:11.640
throw everything to the LEM when in
fact you want to use LMS for what

330
00:23:11.680 --> 00:23:15.839
they're very good at, like pattern
recognition, natural language processing, and you

331
00:23:15.880 --> 00:23:18.720
want to try to still you know, parts of our heavy IQ offering,

332
00:23:19.119 --> 00:23:23.759
which I guess we're kind of jumping
ahead of here, but basically allows users

333
00:23:23.759 --> 00:23:29.599
to ask questions and get answers back
the system. Heavy IQ's writing sequel for

334
00:23:29.640 --> 00:23:33.480
them. It's writing natural answers,
it's picking visualizations. Parts of it,

335
00:23:33.160 --> 00:23:37.920
you know, will actually correct the
SQL queries and figure out, oh,

336
00:23:37.039 --> 00:23:41.519
the user got this little wrong.
We'll do that through traditional software approaches,

337
00:23:41.599 --> 00:23:44.799
right, we won't necessarily trust the
LM to do everything, but use the

338
00:23:44.920 --> 00:23:48.000
LM for the core what it's good
for, which is, yeah, to

339
00:23:48.039 --> 00:23:52.119
take the user's natural language utterance and
translate that into what they actually need or

340
00:23:52.119 --> 00:23:55.759
what they want well. And I
mean, I think that's going to be

341
00:23:55.839 --> 00:24:00.279
a very significant change in the usage
of these tech analogies because when you can

342
00:24:00.319 --> 00:24:03.119
just ask questions and it's not new. I mean, I was joking with

343
00:24:03.160 --> 00:24:07.759
someone the other day. I remember
progress Easy Ask, which was like I

344
00:24:07.799 --> 00:24:11.640
mean twenty years ago, I think
or eighteen years ago, where you would

345
00:24:11.680 --> 00:24:15.160
type in a natural language query and
it would build the sequel for you underneath,

346
00:24:15.160 --> 00:24:18.920
so you could see how it's doing
that. That's not new, but

347
00:24:19.000 --> 00:24:23.480
it's certainly much more sophisticated today than
it was. And to me it's great

348
00:24:23.519 --> 00:24:29.920
because it opens up analysis to a
vastly broader market of business users who don't

349
00:24:29.920 --> 00:24:32.359
know SQL, or don't know SQL
very well, or don't want to sit

350
00:24:32.400 --> 00:24:37.640
around and write sequel. Right,
So that compoundent ended up itself as the

351
00:24:37.759 --> 00:24:41.640
front end is I think, in
a revolutionize and greatly expand Do you see

352
00:24:41.759 --> 00:24:47.279
these technologies and the fact that you're
doing some of the more traditional deterministic stuff

353
00:24:47.359 --> 00:24:52.200
under the covers, I think puts
you in a good category because to your

354
00:24:52.240 --> 00:24:56.119
point, you can ask an LM
all kinds of stuff and you'll get sometimes

355
00:24:56.279 --> 00:24:59.880
very good answers and sometimes very wrong
answers, and you certainly don't want that,

356
00:25:00.720 --> 00:25:03.440
right, So you I think you
have a pretty clever fusion of old

357
00:25:03.480 --> 00:25:08.960
and new technologies in this stack.
And of course heavy is the background,

358
00:25:10.000 --> 00:25:12.279
the dB, that's where a lot
of the action takes place. You're using

359
00:25:12.319 --> 00:25:17.160
the LM on the front end to
facilitate a conversation with that data, but

360
00:25:17.240 --> 00:25:21.799
then you're converting to SQL as part
of your offering, right, so you

361
00:25:22.000 --> 00:25:27.519
are bringing together these two worlds to
be able to enable very very rapid analysis

362
00:25:27.599 --> 00:25:32.599
of very large data sets. Right, Yeah, that's that's correct, And

363
00:25:32.680 --> 00:25:36.880
I think like having the data engine, a very fast form of data engine

364
00:25:37.480 --> 00:25:41.640
is actually dovetails with the use of
the LM. For example, it's great

365
00:25:41.680 --> 00:25:45.519
to have the LM, but part
of what we do to make it accurate

366
00:25:45.880 --> 00:25:49.359
is we actually mine often our customers
have multi billion record tables, so part

367
00:25:49.359 --> 00:25:52.480
of it is that we're actually querying
all that data to figure out what are

368
00:25:52.519 --> 00:25:56.680
the top string values, what's the
time ranges for this data, what are

369
00:25:57.759 --> 00:26:02.319
you know, what are possible and
so we're feeding all that information to the

370
00:26:02.440 --> 00:26:06.000
l M, you know, deterministically
so you can see it. So it's

371
00:26:06.039 --> 00:26:07.480
a great boon to have a very
fast engine so we can pull all that

372
00:26:07.599 --> 00:26:12.440
data dynamically, so you can just
look at tables that the system hasn't seen

373
00:26:12.480 --> 00:26:17.400
before and instantly start being able to
ask questions. Also, like once you

374
00:26:17.519 --> 00:26:19.559
actually generate a query from the LM, you know, it's very nice to

375
00:26:19.599 --> 00:26:23.319
be able to have a fast system
where you want this to be conversational analytics,

376
00:26:23.480 --> 00:26:26.759
not sit and wait analytics. So
used to ask a question, you

377
00:26:26.799 --> 00:26:30.599
know, it could be on a
ten billion record data set. You run

378
00:26:30.680 --> 00:26:33.640
that SQL query, it takes another
you know, half a second or something,

379
00:26:33.960 --> 00:26:36.799
and the user has an answer.
I think. You know, I've

380
00:26:36.839 --> 00:26:40.799
seen some of these other systems and
if they're divorced from the system of record,

381
00:26:40.920 --> 00:26:42.839
if you know, that's you have
your l M service hitting one of

382
00:26:44.160 --> 00:26:47.400
could be terr data, could be
Redshift, could be snowflake. Yeah,

383
00:26:47.440 --> 00:26:51.279
they're not optimizing for their databases,
and so it's it's much less seamless in

384
00:26:51.400 --> 00:26:53.079
terms of being able to get you
know, time to answer, time to

385
00:26:53.160 --> 00:26:57.480
insight. Yeah, so that's interesting
helps a lot. Well, it brings

386
00:26:57.519 --> 00:27:00.279
me to my next big question was
how do you load this thing? So

387
00:27:00.319 --> 00:27:06.519
in order to use this, do
you have to essentially etl in vast amounts

388
00:27:06.559 --> 00:27:08.519
of data? How are you doing
that? Like, what's your mechanism for

389
00:27:08.799 --> 00:27:12.400
populating the database to be able to
get rolling? Yeah, what's cool.

390
00:27:14.000 --> 00:27:18.000
You know what's cool about iav IQ
is you load your data into heavy dB

391
00:27:18.720 --> 00:27:22.720
and you can use our web interface
heavy immerse if you want, you can

392
00:27:22.799 --> 00:27:25.799
do it, you can script it, you can do cover once it's there,

393
00:27:25.839 --> 00:27:29.880
it's automatically variable via natural language via
heavy IQU. You have to do

394
00:27:30.079 --> 00:27:33.440
zero else the system. We have
a kind of server running behind the scenes

395
00:27:33.480 --> 00:27:37.359
that once you start asking questions,
it will know which metadata to pull.

396
00:27:37.680 --> 00:27:41.559
But all of that happens kind of
automatically for the user, and so there's

397
00:27:41.640 --> 00:27:45.759
no setup, no configuration, It
just kind of works out of the box.

398
00:27:45.039 --> 00:27:48.079
The other nice thing is we're running
on GPUs, right, and these

399
00:27:48.359 --> 00:27:52.400
lms run best when accelerated by GPUs, and so we can use that same

400
00:27:52.480 --> 00:27:57.359
infrastructure that we use to run our
database to run the lms themselves, and

401
00:27:57.440 --> 00:28:02.319
so that's kind of cool as well. And then talk about how you would

402
00:28:02.359 --> 00:28:07.160
connect to like a data warehouse for
example, a Snowflake or a Redshift or

403
00:28:07.200 --> 00:28:11.799
something like that. You talked about
being able to use heavy as the front

404
00:28:11.920 --> 00:28:15.440
end basically, so you would be
pulling some data from these warehouses getting it

405
00:28:15.480 --> 00:28:18.160
into this environment. How does that
work? And you know, can you

406
00:28:18.279 --> 00:28:22.920
do things like change data capture for
when there are updates or different things of

407
00:28:22.960 --> 00:28:26.279
this nature. How do you how
do you deal with that sort of threading

408
00:28:26.359 --> 00:28:32.480
of multiple information systems going into heavy. Yeah, that's a great question.

409
00:28:32.599 --> 00:28:36.160
So we have something that we released
a few years ago called heavy connect,

410
00:28:36.759 --> 00:28:41.880
which basically is it's almost a foreign
data wrapper kind of interface for heavy.

411
00:28:41.279 --> 00:28:45.160
So that means we can sit on
top of another store record, whether it's

412
00:28:45.160 --> 00:28:49.920
Snowflake, whether it's Terra Data Redshift
or like a Arcade store or something like

413
00:28:51.000 --> 00:28:55.599
data Bricks, and it seamlessly.
Any queries that you issue will get pushed

414
00:28:55.640 --> 00:28:57.359
down and data and then a data
will get fetched from the underlying store.

415
00:28:59.359 --> 00:29:04.519
We are working on things like change
data capture. Right there's some notion of

416
00:29:04.640 --> 00:29:08.079
it's more built for a pin streams
right now, it can pick up pin

417
00:29:08.119 --> 00:29:15.480
streams updates. You can do retriggers
every night to like fetch the metadata changes.

418
00:29:15.759 --> 00:29:18.519
So we're getting more sophisticated with that. But for most of our customers,

419
00:29:18.880 --> 00:29:21.720
they're not necessarily putting this on top
of their transactional store, and they're

420
00:29:21.759 --> 00:29:26.279
putting on top of another analytics store
in which the underlying data it may be

421
00:29:26.359 --> 00:29:29.960
appended to rapidly, but it's not
changing. You know, you're not seeing

422
00:29:30.000 --> 00:29:33.440
a ton of updates. You may
be seeing deletes, but overall it's a

423
00:29:33.519 --> 00:29:37.880
huge gain because you know, these
systems like Snowflake is a very good system

424
00:29:38.039 --> 00:29:42.839
the separation of computing storage. It's
economical in many ways for users, but

425
00:29:44.000 --> 00:29:48.039
in terms of providing real time analytics. You know, it's less strong and

426
00:29:48.160 --> 00:29:51.799
so using heavy for what it's good
at. They interactive visual analytics, they

427
00:29:51.799 --> 00:29:55.960
interactive quarium and then having that push
down to snowflake is your store record,

428
00:29:56.000 --> 00:29:59.799
I think is really exciting for people. That's pretty cool. Yeah, and

429
00:29:59.880 --> 00:30:04.319
I on your site you talk about
unprecedented context across location and time. That's

430
00:30:04.400 --> 00:30:08.519
kind of where I was going with
earlier in the game. The time series

431
00:30:08.640 --> 00:30:12.960
is so important for understanding what's happening
in any kind of environment. You want

432
00:30:12.960 --> 00:30:15.720
to understand, you know, what
what happened over a period of time,

433
00:30:15.839 --> 00:30:18.920
and what is the key period of
time to understand is it a day,

434
00:30:19.039 --> 00:30:22.640
is it an hour, is it
a week, Sometimes it's a month.

435
00:30:22.079 --> 00:30:27.599
And being able to bounce around in
those different time frames to look at the

436
00:30:27.680 --> 00:30:32.519
difference, look at the impact,
or look at what's what's happening. That's

437
00:30:32.559 --> 00:30:40.400
a really powerful feature to enable analysis
because if you have to reset, you

438
00:30:40.519 --> 00:30:44.519
know, restructure the query, wait
ten minutes or something for it to repopulate

439
00:30:44.599 --> 00:30:48.680
or materialize views or something. Any
any friction you put between the mind of

440
00:30:48.720 --> 00:30:52.759
the user and the information system being
queried is going to slow you down.

441
00:30:52.839 --> 00:30:56.640
And it's going to cause you know, a hindrance and morale to go down,

442
00:30:56.720 --> 00:31:00.759
et cetera. So I think that's
why what you have is so interesting,

443
00:31:00.920 --> 00:31:07.279
because you are greasing the tracks for
the analysts to figure out what is

444
00:31:07.359 --> 00:31:11.559
the appropriate time window I should be
applying to this problem. What do you

445
00:31:11.599 --> 00:31:14.920
think about that? No, I
always say, you know, people talk

446
00:31:14.920 --> 00:31:19.200
about real time analytics, but real
time analytics is often not very powerful without

447
00:31:19.359 --> 00:31:23.920
historical contexts. A good example of
this is one of our longtime customers,

448
00:31:25.039 --> 00:31:30.079
Charter Charter Wireless, Charter Spectrum,
you know, essentially uses us you know,

449
00:31:30.200 --> 00:31:33.799
as they see you know, they
might see an issue, or they

450
00:31:33.880 --> 00:31:37.319
may get something on social media or
people saying I'm having an outage or whatever.

451
00:31:37.720 --> 00:31:41.519
They can instantly go into our system
with a bunch of historical data and

452
00:31:41.640 --> 00:31:45.160
say, okay, what are the
piece of network equipment in that area and

453
00:31:45.279 --> 00:31:48.720
we had faults with them historically?
What's been the network performance this area?

454
00:31:48.839 --> 00:31:53.079
Is this actually an issue or is
this a statistical anomaly? And you know,

455
00:31:53.079 --> 00:31:56.200
I don't think they'd be able to
do that, at least in real

456
00:31:56.319 --> 00:32:00.720
time without without heavy We've seen that
from a lot of our customers. They

457
00:32:00.759 --> 00:32:02.759
say We used to have an outage
or some issue in our system. We

458
00:32:02.839 --> 00:32:07.160
work with a lot of telcos,
and we'd be sitting there, sitting on

459
00:32:07.200 --> 00:32:10.279
our hands waiting for an answer back
from the data engineering team, Versus they

460
00:32:10.319 --> 00:32:14.279
can actually do that themselves. Now
they can dive into the data and often

461
00:32:14.319 --> 00:32:17.319
get to answers in order of magnitude
faster, you know, minutes or tens

462
00:32:17.359 --> 00:32:21.039
of minutes and set of hours or
you know a day or two days,

463
00:32:21.880 --> 00:32:25.440
which really has helped with kind of
faster turnaround times, faster resolution to customer

464
00:32:25.480 --> 00:32:30.480
complaints. And it's all about drilling
in on space and time and being able

465
00:32:30.519 --> 00:32:32.599
to look at the real time data
as well as the historical contexts. Yeah,

466
00:32:32.720 --> 00:32:36.200
I love it. I mean that's
just really really good stuff. So

467
00:32:36.319 --> 00:32:38.799
folks, look at these folks up
online. Heavy dot AI just like it

468
00:32:38.880 --> 00:32:44.440
sounds heavy like the weight and heavy
IQ is the new but heavy dB is

469
00:32:44.480 --> 00:32:45.920
in the background. Todd Mastak,
thank you so much for your time.

470
00:32:46.279 --> 00:32:49.680
Yeah, thanks so much. Or
it's been great. Folks. Stand by.

471
00:32:49.759 --> 00:33:00.799
You're listening to Inside Analysis. Welcome
back to Inside Analysis. Here's your

472
00:33:00.920 --> 00:33:07.480
host, Eric Tabanac. All right, folks, welcome back to Inside Analysis.

473
00:33:07.519 --> 00:33:12.279
Your host here, Eric Kavanaugh,
And next up we have Peter Voss

474
00:33:12.799 --> 00:33:17.319
from a company called the ig Ai
that's a I G O dot ai and

475
00:33:17.720 --> 00:33:22.000
they have a conversational chat but a
chat out with a brain. That's the

476
00:33:22.119 --> 00:33:25.839
tagline. So Peter, welcome to
the show. When you say a chat

477
00:33:25.920 --> 00:33:29.839
out with a brain, what does
that exactly mean? And how is your

478
00:33:29.960 --> 00:33:37.440
chatbot different? Yes, good question. So literally ours has a cognitive engine

479
00:33:37.599 --> 00:33:39.759
or a brain, and as far
as we know, we are the only

480
00:33:39.839 --> 00:33:45.319
company you're offering chatbot with a brain. So what the brain does is it

481
00:33:45.440 --> 00:33:50.279
has short term memory, long term
memory, reasoning, abilowity, and deep

482
00:33:50.440 --> 00:33:55.319
understanding. So it's you know,
it's obviously much more effective than a chatbot

483
00:33:55.359 --> 00:34:00.079
that just relies on a float chart, which is pretty much what everybody we

484
00:34:00.160 --> 00:34:04.880
also doing that you go down a
bloadshot and you know that that's kind of

485
00:34:04.960 --> 00:34:10.039
a decision tree. So we've developed
this over many many years. This is

486
00:34:10.079 --> 00:34:17.199
actually like a third generation of our
technology and allows us to really very effectively

487
00:34:17.840 --> 00:34:23.039
replace calls into agents instead of just
augmenting or instead of just doing you know,

488
00:34:23.159 --> 00:34:29.639
the very simple task. Now,
when you say long term memory and

489
00:34:29.719 --> 00:34:34.400
short term memory, I mean I
understand persisting something either to memory or to

490
00:34:34.559 --> 00:34:38.639
disc for example, but what are
you really talking about? Is there a

491
00:34:38.760 --> 00:34:46.119
semantic layer to this, and then
there's persistent data like concepts and historical context.

492
00:34:46.239 --> 00:34:51.840
Basically go into the details about when
you say memory, what does that

493
00:34:52.000 --> 00:34:55.320
exactly mean and how is that memory
then called and used to create context?

494
00:34:55.639 --> 00:35:02.440
Yes, absolutely, So the backbone
of the system is a dynamic knowledge graph

495
00:35:04.039 --> 00:35:07.039
that basically has the knowledge about the
company, the product, but also the

496
00:35:07.199 --> 00:35:13.559
individual that you're dealing with. So
their personalization is hyper personalization at the individual

497
00:35:13.639 --> 00:35:19.320
level. To complatize it. You
know, one of our customers is one

498
00:35:19.400 --> 00:35:23.719
eight hundred Flowers Harry and David group
of companies, and they use this obviously

499
00:35:23.880 --> 00:35:30.199
as a concercial level service to individual
customers. By the way, just a

500
00:35:30.239 --> 00:35:35.840
few weeks ago in Valentine's Day,
we actually replaced three thousand agents doing that

501
00:35:36.119 --> 00:35:44.480
and you know, very successfully automated
service calls. So what it means is

502
00:35:44.519 --> 00:35:47.599
that the system remembers what you said
in previous conversations. For example, if

503
00:35:47.639 --> 00:35:52.840
you say I want to buy some
chocolate for my niece of birthday, then

504
00:35:52.519 --> 00:35:57.440
that becomes part of the knowledge graph
and that knowledge can then be used so

505
00:35:57.800 --> 00:36:02.199
later on in the conversation you established
what the niece's name is and you know

506
00:36:02.360 --> 00:36:08.360
her address or whatever. That information
is kept in the brain and it's available

507
00:36:08.480 --> 00:36:15.280
for future conversations or to follow up
conversations. So the long term memory is

508
00:36:15.320 --> 00:36:22.320
an integral part of the knowledge graph, and that is then used both for

509
00:36:22.440 --> 00:36:28.480
short term memory as appropriate and for
context during the conversations. As you say,

510
00:36:28.519 --> 00:36:34.920
it's a semantic level understanding. Well, that's interesting. So basically,

511
00:36:35.079 --> 00:36:38.599
every time there is a customer call
of guessing, you're creating a new node

512
00:36:39.440 --> 00:36:45.159
and new edges on that knowledge graph, and it's ear marked as a node

513
00:36:45.320 --> 00:36:50.559
of this conversation I'm having with John
Smith, for example, on this date.

514
00:36:50.679 --> 00:36:52.840
Is that correct? So every conversation. Yeah, And actually it's a

515
00:36:52.880 --> 00:37:00.800
little bit more sophisticated than that,
because you really want to separate out individual

516
00:37:00.880 --> 00:37:04.519
customers. You don't want to you
know, combine information that you have any

517
00:37:04.639 --> 00:37:09.960
risk of, you know, you
know, security, security risk. So

518
00:37:10.079 --> 00:37:15.719
what we do is actually there's a
three level architecture of the knowledge graph.

519
00:37:15.800 --> 00:37:21.599
The the core level, the inner
level is information shared by all of our

520
00:37:21.679 --> 00:37:24.559
customers, so it's just you know, the generic information. The next level

521
00:37:24.639 --> 00:37:30.519
out is the unique information for the
enterprise that we're working with, like their

522
00:37:30.599 --> 00:37:35.719
business rules, you know their product
ontology and everything you need to know about

523
00:37:35.760 --> 00:37:39.400
the business. And then the third
layer is unique for each individual customer.

524
00:37:39.400 --> 00:37:45.719
In the case of one eight hundred
Flowers, it's twenty million customers that each

525
00:37:45.800 --> 00:37:49.760
have their own part of the knowledge
graph that is loaded. As soon as

526
00:37:49.760 --> 00:37:53.559
we identify with the customer is we
load that additional part of the knowledge graph.

527
00:37:53.599 --> 00:37:58.599
And then when the conversation finishes with
whatever has been added that is then

528
00:37:59.480 --> 00:38:05.840
you know it away. That's very
interesting. So is the is phone number

529
00:38:06.679 --> 00:38:12.599
the key value to identify who's who? Or is it phone and some other

530
00:38:13.079 --> 00:38:17.800
Oh? It really it really depends
on what the customer provides and what channel

531
00:38:17.880 --> 00:38:22.480
you're using, because we you know, our system is truly omni channel,

532
00:38:22.519 --> 00:38:27.480
so you can really start a conversation
from a website. You can go to

533
00:38:27.639 --> 00:38:30.760
chat, but then you could switch
over to phone. So if you're coming

534
00:38:30.800 --> 00:38:36.039
through chat, might well be emailed
address, or it might be the order

535
00:38:36.159 --> 00:38:38.400
number that you give to identify the
customer, or it could be the phone

536
00:38:38.480 --> 00:38:43.360
number. Of course, there has
to be validation to make sure you are

537
00:38:43.719 --> 00:38:47.039
in fact talking to the right person. But the identification really, you know,

538
00:38:47.199 --> 00:38:51.400
can happen through different things. Now, if it's coming in through say

539
00:38:51.440 --> 00:38:55.880
Apple Business Chat, you already have
validation of the customer. It's coming from

540
00:38:55.920 --> 00:39:00.719
a web website where the customer is
already logged in, then you'll you'll actually

541
00:39:00.760 --> 00:39:07.320
have a customer ID. That's interesting
and so hmm, so you're replacing call

542
00:39:07.519 --> 00:39:12.599
centers. Basically, are you able
to do voice or is it only chat?

543
00:39:13.159 --> 00:39:16.239
We do we do both voice and
chat. Yeah, it's truly omni

544
00:39:16.360 --> 00:39:22.079
channel that you can. You can
because it's the same brain whether you're switching

545
00:39:22.119 --> 00:39:27.440
your channel, your communication channel,
whether that switches from you know, SMS

546
00:39:27.599 --> 00:39:31.760
to Apple Business Chat to voice,
it's the same brain and you'll have the

547
00:39:31.840 --> 00:39:37.280
same context that available. So is
it a computerized voice, then it's just

548
00:39:37.360 --> 00:39:45.760
a computer voice that's having it's correct
for the lephany as computerized text to speech,

549
00:39:45.239 --> 00:39:50.519
you know those does that become really
quite good these days. And I

550
00:39:50.559 --> 00:39:54.079
think one of the important things is
we believe it's important that you let the

551
00:39:54.199 --> 00:40:00.599
customer know that it isn't AI they're
talking to I. Really you don't like

552
00:40:00.679 --> 00:40:05.079
it when enterprise want to full customers
into you know, it's got to sound

553
00:40:05.199 --> 00:40:07.679
just like a human. I don't
think that's a good idea at all.

554
00:40:08.440 --> 00:40:14.119
We always give a well subject to
the enterprise we're working with, but we

555
00:40:14.360 --> 00:40:17.920
very much want to design the system
in a way that the customer can ask

556
00:40:19.079 --> 00:40:23.639
for a live operator at any time. Very simple. There's no you know,

557
00:40:23.679 --> 00:40:28.159
there should be no barrier to talking
to a human if you want to.

558
00:40:28.400 --> 00:40:31.199
For whatever reason. You know,
either the system doesn't have the knowledge

559
00:40:31.320 --> 00:40:37.039
or somebody to simply simply prefers it. But you know, in the example

560
00:40:37.079 --> 00:40:43.280
I just gave them a Valentine's date, ninety percent of the interactions were handled

561
00:40:43.519 --> 00:40:50.239
entirely through automation by our system,
and only you know, ten percent opted

562
00:40:50.440 --> 00:40:55.199
or needed to go to an operator. That's pretty interesting, So you do

563
00:40:55.440 --> 00:40:58.719
inform people. It's funny because that
was going to be a question I asked

564
00:40:58.760 --> 00:41:01.199
you, Is that you telling people
that this is a computerized chat butt and

565
00:41:01.320 --> 00:41:06.000
you are have you seen have you
seen much pushback on that? Or are

566
00:41:06.039 --> 00:41:08.519
most people okay with that? Oh? No, not not at all.

567
00:41:08.639 --> 00:41:13.599
You know, I was recently at
a conference and and and somebody said to

568
00:41:13.719 --> 00:41:17.320
me, we're talking about the difference
between chat and phone, And he's saying,

569
00:41:17.480 --> 00:41:21.320
you know, surely boomers, you
know, they still want to talk

570
00:41:21.400 --> 00:41:24.159
by by phone. They don't want
to use chat. And I looked at

571
00:41:24.199 --> 00:41:29.599
him and I said, well,
do you prefer talking to somebody? Do

572
00:41:29.679 --> 00:41:32.760
you prefer to use chat? He
said, well chat? Well, you're

573
00:41:32.760 --> 00:41:37.880
a boomer on't you. So I
think people are increasingly becoming comfortable with with

574
00:41:38.119 --> 00:41:44.280
smartphones. Everybody is, and people
actually learning that it's that it's it's easier

575
00:41:44.360 --> 00:41:49.119
to text in many cases. And
of course the younger generation doesn't even want

576
00:41:49.199 --> 00:41:52.800
to talk to to to a person. People just want to get stuff done.

577
00:41:53.079 --> 00:41:58.000
That's really what we're finding. As
long as your automation, whether it's

578
00:41:58.119 --> 00:42:04.639
voice or text, is moving forward, and you know that people know this

579
00:42:04.800 --> 00:42:07.159
thing is really helping me. You
know, I'm getting things done. They

580
00:42:07.199 --> 00:42:09.960
don't want they don't want to talk
to humans. I just want to get

581
00:42:10.000 --> 00:42:15.079
things done. That's that's really what
we're finding. But you know, as

582
00:42:15.119 --> 00:42:19.159
soon as it becomes frustrating, you
know, press one, please listen carefully.

583
00:42:19.400 --> 00:42:22.239
Options have recently changed. You know, all your business is important to

584
00:42:22.400 --> 00:42:24.559
us. I mean, you know, you pull your hair off and you

585
00:42:24.719 --> 00:42:30.559
just press zero zero zero, get
me. It's an operator. The society

586
00:42:30.079 --> 00:42:34.480
has to be useful, you know, and it has to get you to

587
00:42:34.960 --> 00:42:37.920
what you want to do quickly.
Yeah, you know, we're going to

588
00:42:37.960 --> 00:42:40.559
do the podcast bonus seven here in
just a moment, and I think we'll

589
00:42:40.599 --> 00:42:47.119
dive into how you enrich or how
you fuel the chatbot. I'm guessing you

590
00:42:47.320 --> 00:42:52.800
use corporate documentation, a history of
the conversations that people have had, things

591
00:42:52.840 --> 00:42:55.119
of that nature, because you have
to have a starting point, and then

592
00:42:55.159 --> 00:43:00.400
of course over time, you curate
and you refine and you find in tune

593
00:43:00.440 --> 00:43:02.280
and you get these things to where
they're pretty good at being able to handle

594
00:43:02.360 --> 00:43:07.400
whatever questions the person might ask.
And that is the key, right because

595
00:43:07.440 --> 00:43:10.000
customer service is very important. Everyone
says customer experience. It's all about the

596
00:43:10.119 --> 00:43:15.840
experience. It's going to be mindful
of all that while still leveraging modern technology.

597
00:43:15.880 --> 00:43:19.320
Well, folks were talking to Peter
Voss of Igo dot Ai. Be

598
00:43:19.480 --> 00:43:25.039
right back for the podcast bonus segment. Standby, Well, okay, folks,

599
00:43:25.079 --> 00:43:29.920
time for the podcast bonus segment here
on Inside Analysis talking to Peter Voss

600
00:43:30.119 --> 00:43:35.039
of Igo dot ai, and we're
talking about your chat by your conversational chatbot

601
00:43:35.079 --> 00:43:37.960
with a brain. I love that
it's got long term memory, it's got

602
00:43:37.000 --> 00:43:39.960
a knowledge graph. I'm a big
fan of knowledge graphs. I think they're

603
00:43:40.000 --> 00:43:45.159
going to be very helpful in the
AI age because not just do they store

604
00:43:45.199 --> 00:43:51.239
information, but they store it in
a way it's very easy to access in

605
00:43:51.440 --> 00:43:54.960
terms of relevance, because the relevance
are on the edges, right, You

606
00:43:55.000 --> 00:44:00.119
have nodes and edges in a graph, nodes of the entities, edges of

607
00:44:00.159 --> 00:44:04.320
the characteristics basically, and so in
a knowledge graph you can capture a lot

608
00:44:04.360 --> 00:44:08.559
of information about individuals, their buying
habits, the characteristics of who they are.

609
00:44:08.719 --> 00:44:13.360
Those are all the edges. First
of all, is that true with

610
00:44:13.559 --> 00:44:17.280
your solution? And then second of
all, tell us how you actually populate

611
00:44:17.719 --> 00:44:22.719
the knowledge graph in the first place
to get rolling. What sort of information

612
00:44:22.239 --> 00:44:25.480
can you load into this graph?
And in what format do you load it

613
00:44:25.559 --> 00:44:30.239
to be able to get somewhere?
Yeah, this is exactly right. And

614
00:44:31.079 --> 00:44:37.119
you know, an important benefit of
the approach of a knowledge growth based system

615
00:44:37.760 --> 00:44:44.320
is that it's truly scrutable and it's
auditable and as opposed to know the current

616
00:44:44.360 --> 00:44:47.599
trend where everybody's trying to do things
with statistical models, large language models,

617
00:44:49.239 --> 00:44:53.199
which are you know not they black
boxes, they can't be ordered it and

618
00:44:53.280 --> 00:44:59.039
they hallucinate. You don't have that
with us. Dappa actually calls this the

619
00:44:59.199 --> 00:45:02.960
third wave of AI, where you
have cognitive AI, a cognitive system as

620
00:45:04.000 --> 00:45:09.119
opposed to the statistical system. So
the way we populate the knowledge graph is,

621
00:45:09.280 --> 00:45:13.039
you know, as I mentioned in
the previous section, there are three

622
00:45:13.119 --> 00:45:16.840
layers. The middle layer, the
center layer, is the common knowledge that

623
00:45:17.360 --> 00:45:22.280
is required for every conversation. You
know about people, places, time,

624
00:45:22.559 --> 00:45:28.760
and you know, the common knowledge
and that knowledge we curated over several years

625
00:45:29.280 --> 00:45:34.599
to act as the core of the
intelligence of the conversational engine, and we

626
00:45:34.719 --> 00:45:38.559
keep adding to that. But then
the second layer is basically where we import

627
00:45:39.079 --> 00:45:46.719
the customers ontologies for their different products
that they have product categories and the various

628
00:45:46.800 --> 00:45:51.639
business rules and so on. So
that is then added to the knowledge graph

629
00:45:51.760 --> 00:45:57.559
and you know, just carefully curated
to make sure that you have accurate information

630
00:45:59.280 --> 00:46:04.480
in that little lay. You also
have connections via APIs to the back end

631
00:46:04.559 --> 00:46:07.559
system to obviously give you like you
know, the current status of an order

632
00:46:07.760 --> 00:46:14.719
and to be able to get the
basic information from the back end system.

633
00:46:15.000 --> 00:46:21.400
And then the ult layer is the
one that is unique to each individual end

634
00:46:21.519 --> 00:46:29.719
user customer, and that is built
up dynamically through the conversation by basically expecting

635
00:46:29.760 --> 00:46:35.159
the semantics of the conversation and adding
that to the knowledge graph as you go

636
00:46:35.199 --> 00:46:39.079
along. And you know, as
you pointed out, the obviously getting all

637
00:46:39.119 --> 00:46:46.280
the business rules and interpretations and the
ontologies and the specialized language that people use

638
00:46:46.440 --> 00:46:51.840
or how people might ask for help, all of that definitely requires tuning.

639
00:46:52.039 --> 00:46:58.119
I mean, it doesn't matter how
practiced you are at this art, you

640
00:46:58.599 --> 00:47:02.400
never you can never fully anticipate how
people will actually use the system, So

641
00:47:02.840 --> 00:47:08.079
the ongoing tuning is definitely important.
Also from the point of view that you

642
00:47:08.159 --> 00:47:14.960
know, products and business rules and
demographics change all the time in a company,

643
00:47:15.480 --> 00:47:20.320
so you don't want the system to
deteriorate as it gets out of sync

644
00:47:20.840 --> 00:47:25.280
with whatever you know. The company
and customer dynamic is. I guess that's

645
00:47:25.320 --> 00:47:30.599
my last question to you, which
is, you know, the world changes,

646
00:47:30.840 --> 00:47:37.519
habits change, products change, but
it's hard for a model to unlearn

647
00:47:37.800 --> 00:47:47.119
something. So in your graph,
do you have the capacity to update offerings

648
00:47:47.239 --> 00:47:52.239
for example, like what the chatbot
is saying. Obviously it's pulling from the

649
00:47:52.320 --> 00:47:54.920
knowledge graph to some extent and you
can curate that and find tune it,

650
00:47:54.960 --> 00:47:59.880
et cetera. But can you actually
go in and overwrite. Can you go

651
00:48:00.159 --> 00:48:04.000
in and say, Okay, this
old product is not in our service anymore.

652
00:48:04.440 --> 00:48:06.360
Here's a new product. And so
how do you do that? In

653
00:48:06.400 --> 00:48:10.159
other words, how do you maintain
the repository of information such that it's all

654
00:48:10.360 --> 00:48:15.880
current and you don't have all the
information sitting around that is going to throw

655
00:48:15.960 --> 00:48:22.880
people off, right absolutely. And
the big difference with a cognitive AI approach

656
00:48:22.960 --> 00:48:28.559
with the knowledge croft approach that we
have is you can literally retrain the system

657
00:48:28.679 --> 00:48:35.679
from scratch in minutes, and you
know, and it costs next to nothing.

658
00:48:35.880 --> 00:48:39.000
It's not the one hundred million dollar
type of model that you know,

659
00:48:39.199 --> 00:48:43.800
large language models, where you have
to retrain the whole model and you really

660
00:48:43.920 --> 00:48:47.119
cannot change the information in the model. We don't have that problem at all.

661
00:48:47.599 --> 00:48:52.079
So you can increment the change knowledge, or you can retrain the model

662
00:48:52.119 --> 00:48:59.039
if bigger changes happen from scratch,
and literally it will train it in minutes.

663
00:48:59.199 --> 00:49:04.679
That's how effect that this kind of
approach is. Because it's not working

664
00:49:04.760 --> 00:49:07.760
with ten trollium bits of information,
right yeah, right, that's because it's

665
00:49:07.760 --> 00:49:12.559
a smaller model, right, I
mean it's would you call it a small

666
00:49:12.679 --> 00:49:15.719
language model or it's just a knowledge
graph basically the knowledge graph, I think

667
00:49:15.840 --> 00:49:22.639
is a better way to describe it. The language capability are also part of

668
00:49:22.719 --> 00:49:28.960
the knowledge part part of the knowledge
graph. But you know that part has

669
00:49:29.039 --> 00:49:32.800
been built of a long time.
And I mean basic language understanding doesn't doesn't

670
00:49:32.840 --> 00:49:37.360
really change. You may add extra
synonyms, you know, or new product

671
00:49:37.480 --> 00:49:42.119
names and so on, that they're
part of the ontology. The core language

672
00:49:42.239 --> 00:49:47.519
understanding doesn't really doesn't really change.
Gotcha, very very cool stuff. So

673
00:49:49.639 --> 00:49:52.480
I go dot AI. You were
conversational AI. So you work with one

674
00:49:52.719 --> 00:49:57.480
hundred flowers. Basically anyone who has
a call center, that's a good use

675
00:49:57.519 --> 00:50:01.440
case for you, right, yes, call center replacing call center agents really

676
00:50:01.639 --> 00:50:07.639
is the most obvious use case for
us, but can also be used as,

677
00:50:07.880 --> 00:50:10.880
for example, a diabetes coach,
you know, a hyper personalized diabetes

678
00:50:10.920 --> 00:50:16.360
coach that'll ask you and learn whether
you love broccoli or you know you're vegan,

679
00:50:16.559 --> 00:50:20.960
or what you know, whether you
when you want to do exercise,

680
00:50:21.039 --> 00:50:23.159
if you if you want to exercise, and what kind of thing well,

681
00:50:24.159 --> 00:50:30.840
or an assistant for salespeople. So
it has many different use cases. The

682
00:50:30.960 --> 00:50:36.199
core engine that we have, the
cognitive AI engine. But our right now,

683
00:50:36.280 --> 00:50:40.400
our focus and our strength really is
in replacing call center agents very effectively.

684
00:50:43.119 --> 00:50:45.199
Good stuff. Well, look these
folks up online, Peter Voss,

685
00:50:45.320 --> 00:50:50.119
Theoss from Igo. That's a I
T O dot AI. Thanks for your

686
00:50:50.239 --> 00:50:52.719
time and folks want to be in
the show. Sending email in for at

687
00:50:52.760 --> 00:50:58.360
inside analysis dot com. You've been
listening to Inside Analysis and now the voices

688
00:50:58.440 --> 00:51:04.039
of GACAA for an exciting announcement.
Want to hear NBC and News or KCAA

689
00:51:04.239 --> 00:51:07.480
anywhere you go, Well, now
there's an app for that. CACAA is

690
00:51:07.519 --> 00:51:10.840
celebrating twenty five years in our silver
anniversary with a brand new app. The

691
00:51:10.960 --> 00:51:15.920
new KCAA app is now available on
your smart device, cell phone, in

692
00:51:16.000 --> 00:51:21.440
your car or any place. Just
search KCAA on Google Play or in the

693
00:51:21.480 --> 00:51:24.440
Apple Store one touch and you can
listen on your car radio, Bluetooth device,

694
00:51:24.599 --> 00:51:30.239
Android Auto or Apple Car Play,
catch the KCAA buzz in your earbuds

695
00:51:30.400 --> 00:51:34.360
or on the streets. Celebrating twenty
five years of talk news and excellence with

696
00:51:34.480 --> 00:51:39.320
our new KCAA app. Just do
it and download it. KCAA celebrating twenty

697
00:51:39.400 --> 00:51:47.119
five year to Heblt Club's original pure
powdy Rco super TA helps build red corpuscles

698
00:51:47.159 --> 00:51:52.000
in the blood, which carry oxygen
too our organs and cells. Our organs

699
00:51:52.039 --> 00:51:57.039
and cells need oxygen to regenerate themselves. The immune system needs oxygen to develop,

700
00:51:57.079 --> 00:52:00.920
and cancer dies in oxygen. So
the T is great for healthy people

701
00:52:00.000 --> 00:52:04.679
because it helps build the immune system, and it can truly be miraculous for

702
00:52:04.840 --> 00:52:08.320
someone fighting a potentially life threatening disease
due to an infection, diabetes, or

703
00:52:08.519 --> 00:52:13.719
cancer. The T is also organic
and naturally caffeine free. A one pound

704
00:52:13.760 --> 00:52:16.400
package of T is forty nine ninety
five, which includes shipping. To order,

705
00:52:16.480 --> 00:52:21.480
please visit to Hebot club dot com. To hebo is spelled T like

706
00:52:21.599 --> 00:52:25.199
tom, a h ee b like
boy oh. Then continue with the word

707
00:52:25.280 --> 00:52:30.480
T and then the word club.
The complete website is to Hebot Club dot

708
00:52:30.599 --> 00:52:34.679
com or call us at eight one
eight sixty one zero eight zero eight eight

709
00:52:34.880 --> 00:52:38.199
Monday through Saturday nine am to five
pm California time. That's eight one eight

710
00:52:38.280 --> 00:52:44.599
sixty one zero eight zero eight eight
to Ebot club dot com. With sixty

711
00:52:44.840 --> 00:52:51.840
years of fascinating facts. This is
the man from Yesterday and back in time

712
00:52:51.920 --> 00:52:54.840
we go to this time in nineteen
sixty four, the Beatles have six songs

713
00:52:54.920 --> 00:52:59.239
in the top thirty, including do
You Want to Know a Secret. Another

714
00:52:59.280 --> 00:53:01.559
British band is breaking through. They're
called the Searchers. That band got its

715
00:53:01.599 --> 00:53:06.159
name from a John Winn movie.
Here is Sugar and Spice, Sugar and

716
00:53:06.360 --> 00:53:19.360
Spier gives a se Sugar Spies you
know the Licus And from this time in

717
00:53:19.480 --> 00:53:22.920
nineteen seventy four, looks like NBC
is adding The Rockford File starring James Garner,

718
00:53:23.039 --> 00:53:27.239
from the world premiere movie shown just
a few weeks ago. Also being

719
00:53:27.280 --> 00:53:30.239
added Little House on the Prairie starring
Michael Landon from the recent high rated TV

720
00:53:30.400 --> 00:53:34.599
movie This is Jim Rockford. It
don't leave your name and message. I'll

721
00:53:34.639 --> 00:53:37.719
get back to you. Hi,
Jim, thanks for the dinner invitation.

722
00:53:37.400 --> 00:53:40.639
I'd love to, but does it
have to be the taco stand? And

723
00:53:40.719 --> 00:53:45.840
from this time in nineteen fifty seven, NBC TV says it is experimenting with

724
00:53:45.960 --> 00:53:50.199
the new videotape on some of its
more important shows, among them The Steve

725
00:53:50.239 --> 00:53:54.280
Allen Show, Caesar's Hour starring Sid
Caesar, and The Sunday Night Donna Shore

726
00:53:54.480 --> 00:54:00.880
Chevy Show Chevrolet in behalf of your
local authorized Chevrolet Dealer Live from Mollywood.

727
00:54:01.119 --> 00:54:10.599
God Dinah Shore Heavy Show with more
at Man from yesterday dot com. It's

728
00:54:10.679 --> 00:54:15.039
that time of year again, No, not the holidays. Medicare open enrollment

729
00:54:15.280 --> 00:54:20.199
and if you have questions about Medicare, you should talk to the local experts.

730
00:54:20.320 --> 00:54:23.360
Paul Barrette and Associates, all of
his agents are certified with plans that

731
00:54:23.400 --> 00:54:28.119
are accepted by most of the medical
groups in our area. Call nine oh

732
00:54:28.239 --> 00:54:32.239
nine seven nine three oh three eight
five. Their services free and after forty

733
00:54:32.280 --> 00:54:36.559
two years of the business, their
agents are trained to help you pick the

734
00:54:36.639 --> 00:54:40.760
plan that's right for you. KCAA
Radio has openings for one hour talk shows.

735
00:54:40.880 --> 00:54:44.440
If you want to host a radio
show, now is the time.

736
00:54:44.760 --> 00:54:49.599
Make KCIA your flangship station. Our
rates are affordable and our services are second

737
00:54:49.679 --> 00:54:52.760
to none. We broadcast to a
population of five million people plus. We

738
00:54:52.880 --> 00:54:58.239
stream and podcast on all major online
audio and video systems. If you've been

739
00:54:58.280 --> 00:55:02.840
thinking about broadcasting a week radio program
on real radio plus the internet, contact

740
00:55:02.880 --> 00:55:07.719
our CEO at two eight one five
nine nine ninety eight hundred two eight one

741
00:55:07.960 --> 00:55:12.599
five nine nine ninety eight hundred.
You could skype your show from your home

742
00:55:12.679 --> 00:55:16.119
to our Redlands, California studio,
where our live producers and engineers are ready

743
00:55:16.159 --> 00:55:21.880
to work with you personally. A
radio program on KCIA is the perfect work

744
00:55:21.920 --> 00:55:27.079
from home advocation in these stressful times. Just type KCA radio dot com into

745
00:55:27.119 --> 00:55:30.679
your browser to learn more about hosting
a show on the best station in the

746
00:55:30.800 --> 00:55:35.639
nation, or call our CEO for
details to eight one five nine nine ninety

747
00:55:35.639 --> 00:55:40.239
eight hundred. Ralph Waldo Emerson told
of a dinner guest who went on and

748
00:55:40.360 --> 00:55:45.400
on about the virtue of honesty,
offering his own life as a model of

749
00:55:45.480 --> 00:55:49.800
perfect rectitude. The louder he talked
of his honors, said Emerson, the

750
00:55:49.920 --> 00:55:54.559
faster we counted our spoons. That's
my reaction to the cacophony of phony piety

751
00:55:54.639 --> 00:56:00.800
arising from Republican governors and legislators who
are trying to enact more than two hundred

752
00:56:00.800 --> 00:56:04.800
and fifty new state laws to stop
black, Latino, Asian, American,

753
00:56:05.000 --> 00:56:09.840
Indigenous, and other non Caucasian voters
from casting ballots. Yet they proclaim we're

754
00:56:09.880 --> 00:56:15.000
not racists, we're righteous crusaders protecting
the sanctity of the vote. Really,

755
00:56:15.440 --> 00:56:21.039
so, why are they specifically targeting
people of color with their repressive voting restrictions.

756
00:56:21.599 --> 00:56:27.119
For example, panicky Republican lawmakers in
Georgia tried to outlaw any early voting

757
00:56:27.239 --> 00:56:32.400
on Sundays. Odd why it's a
flagrantly racist attack on the black church.

758
00:56:34.079 --> 00:56:38.079
For years, a joyous civic tradition
called Souls to the Poles has played out

759
00:56:38.159 --> 00:56:43.800
in southern Black churches on Sundays prior
to election Day. After the sermon and

760
00:56:43.880 --> 00:56:47.639
prayers, congruentts, ministers, musicians, and others in the church family travel

761
00:56:47.679 --> 00:56:52.519
in a caravan to early voting locations
to cast ballots. It turns voting into

762
00:56:52.599 --> 00:56:58.679
a civic, spiritual, and fun
experience. What kind of shriveled soul tries

763
00:56:58.760 --> 00:57:04.400
to kill that? Apparently the same
shameful souls and the Georgia GOP who want

764
00:57:04.440 --> 00:57:08.320
to stop local groups from providing water
and snacks to citizens forced to wait for

765
00:57:08.519 --> 00:57:13.639
hours in line to vote. They're
actually trying to make it a crime to

766
00:57:13.760 --> 00:57:17.960
give water to thirsty voters. Hey, Republicans, what would Jesus do?

767
00:57:19.960 --> 00:57:23.639
This is Jim Heitr Saying the goal
and duty of every public official ought to

768
00:57:23.719 --> 00:57:28.880
be to maximize voter turnout. After
all, the more Americans who vote,

769
00:57:29.039 --> 00:57:34.440
the stronger are democracy. But there's
the ugly political truth. Republican officials no

770
00:57:34.559 --> 00:57:39.480
longer support democracy. KILLISTINAKCAA Lowell,
Linda at one six point five FM,

771
00:57:39.679 --> 00:57:45.519
K two ninety three, c F
Marino Valley, NBC News Radio. I'm

772
00:57:45.599 --> 00:57:51.360
Chris Kragio. The US is rejecting
claims by Russian President Vladimir Putin that Ukraine

773
00:57:51.440 --> 00:57:54.000
may have been involved in a deadly
attack on a Moscow concert hall. In

774
00:57:54.039 --> 00:57:58.760
an ABC interview on Sunday, Vice
President Kamala Harris said there is no evidence

775
00:57:58.800 --> 00:58:01.960
of that whatsoever. He added that
we know isis k is actually responsible for

776
00:58:02.159 --> 00:58:07.639
what happened. Former Republican National Committee
Chairerona McDaniel says she disagrees with Donald Trump's

777
00:58:07.639 --> 00:58:12.960
promise to free individuals involved in the
attack on the US Capitol. If you

778
00:58:13.119 --> 00:58:16.119
attacked our capital and you have been
you and you've been convicted, then that

779
00:58:16.199 --> 00:58:20.440
should stay. During an interview on
NBC's Meet the Press, McDaniel said the

780
00:58:20.519 --> 00:58:23.880
violence that happened on January sixth was
unacceptable. She said those convicted should serve

781
00:58:23.960 --> 00:58:29.000
out their sentence. McDaniel emphasized,
however, she doesn't hold the former president

782
00:58:29.079 --> 00:58:32.159
responsible for the riot. She stepped
down as chairwoman of the RNC earlier this

783
00:58:32.320 --> 00:58:37.840
month and will join NBC News as
a political analyst. The interview took place

784
00:58:37.920 --> 00:58:42.800
prior to NBC News announcing McDaniel's hiring. Former Supreme Court Justice Stephen Bryer is

785
00:58:42.880 --> 00:58:46.199
speaking about the High Court's decision to
overturn roeb Wade and what he and two

786
00:58:46.280 --> 00:58:50.880
other justices wrote in their dissent.
One of the things we said is what

787
00:58:51.039 --> 00:58:53.639
we fear. They think this will
be simpler the majority, because it will

788
00:58:53.719 --> 00:58:57.719
leave it all up to the states. We don't think it will be simpler.

789
00:58:58.039 --> 00:59:00.639
We think that there will be a
lot of more cases coming up,

790
00:59:00.840 --> 00:59:04.440
Bryar told NBC's Meet the Press the
leak of the decision was unfortunate. He

791
00:59:04.559 --> 00:59:07.400
said he had a theory about who
leaked the decision, but declined to share

792
00:59:07.440 --> 00:59:09.719
any names. The former justice added, however, he'd be amazed if it

793
00:59:09.880 --> 00:59:15.199
was a judge. Briar said it's
possible the Supreme Court could one day overrule

794
00:59:15.320 --> 00:59:19.320
the decision. La Dodger starshow heey
Otani is set to address the media tomorrow

795
00:59:19.400 --> 00:59:22.840
for the first time since allegations of
illegal gambling and theft involving his former interpreter's

796
00:59:22.880 --> 00:59:28.559
surface. Major League Baseball says it
has started a formal investigation. The interpreter

797
00:59:28.719 --> 00:59:31.360
is accused of receiving more than four
and a half million dollars from Otani to

798
00:59:31.480 --> 00:59:36.840
payoff gambling debts made with an illegal
bookie. The Dodgers fired the interpreter after

799
00:59:36.920 --> 00:59:39.880
The La Times reported on the case. I'm Chris Caragio, NBC News Radio,

800
00:59:43.119 --> 00:59:49.000
NBC News on CACAA Lomel sponsored by
Teamsters Local nineteen thirty two Protecting the

801
00:59:49.079 --> 01:00:05.159
Future of Working Families, Teamsters nineteen
thirty two dot org. Was your car

802
01:00:05.280 --> 01:00:08.360
involved in an accident or just need
help with dents? All Magic Paint and by