1
00:00:07,119 -->
00:00:10,030
Gordon Wong: Welcome to Hard Problems,
Smart Solutions, the Newfire Podcast,
2
00:00:10,079 -->
00:00:12,640
where we explore the most complex
challenges and groundbreaking
3
00:00:12,640 -->
00:00:14,099
solutions with industry leaders.
4
00:00:14,370 -->
00:00:17,750
I'm Gordon Wong, VP of Data
AI at Newfire Global Partners,
5
00:00:18,140 -->
00:00:19,159
and your host for this episode.
6
00:00:21,823 -->
00:00:24,423
Today, I'm thrilled to welcome
Sandeep Dhamale, Director of
7
00:00:24,443 -->
00:00:27,554
Engineering, Data and Intelligence
at the American Medical Association.
8
00:00:28,143 -->
00:00:31,233
Sandeep has built his career leading
data and platform engineering teams.
9
00:00:31,653 -->
00:00:36,044
At the AMA, he has spearheaded
transformative projects like Datalabs
10
00:00:36,074 -->
00:00:38,593
GPT, a private LLM infrastructure.
11
00:00:39,178 -->
00:00:42,848
and the AMA Intelligent Platform, which
brings a modern, reusable technical
12
00:00:42,848 -->
00:00:45,898
infrastructure to AMA's data initiatives.
13
00:00:46,158 -->
00:00:50,759
Beyond his roles at AMA, Sandeep
has also served as an advisor with
14
00:00:50,808 -->
00:00:53,578
Newfire, bringing his expertise
to help drive innovation and
15
00:00:53,578 -->
00:00:54,968
scalability for our clients.
16
00:00:55,368 -->
00:00:58,498
Today, we'll explore how healthcare
organizations can adopt scalable
17
00:00:58,498 -->
00:01:01,718
data solutions to improve
patient and operational outcomes.
18
00:01:01,938 -->
00:01:03,548
Sandeep, welcome to the podcast.
19
00:01:04,809 -->
00:01:05,488
Sandeep Dhamale: Thank you, Gordon.
20
00:01:05,548 -->
00:01:08,969
Really excited to be here and I
look forward to a fun conversation.
21
00:01:09,538 -->
00:01:10,058
Gordon Wong: Yeah, me too.
22
00:01:10,058 -->
00:01:12,698
This is, these are typically a lot
of fun and I've been looking forward
23
00:01:12,698 -->
00:01:13,768
to speaking to you for a while.
24
00:01:14,488 -->
00:01:16,719
One of the things I wanted to
ask you about really is how
25
00:01:16,719 -->
00:01:18,339
you got into this space, right?
26
00:01:18,339 -->
00:01:19,578
You were originally in fintech.
27
00:01:19,579 -->
00:01:22,868
Can you talk a bit about how
your career journey took you from
28
00:01:22,868 -->
00:01:26,489
fintech to healthcare and how you
can, how your fintech experience
29
00:01:26,519 -->
00:01:29,609
shaped your strategies for tackling
healthcare's unique data challenges?
30
00:01:30,547 -->
00:01:31,517
Sandeep Dhamale: Yeah, great.
31
00:01:32,257 -->
00:01:37,177
I mean, I consider myself very lucky to
have had a career journey that I had.
32
00:01:37,637 -->
00:01:41,847
Basically, after I finished my master's in
computer science, right, I landed this job
33
00:01:41,877 -->
00:01:44,177
at a global fintech firm called SunGard.
34
00:01:44,687 -->
00:01:46,617
The job was great for my career.
35
00:01:46,617 -->
00:01:51,497
It allowed me to expand my experience way
beyond software engineering skills, right?
36
00:01:51,497 -->
00:01:55,357
It was a global development firm
with an emphasis on building
37
00:01:55,817 -->
00:01:57,267
global market connectivity.
38
00:01:57,277 -->
00:02:00,547
So I was really talking to
different various global exchanges.
39
00:02:00,547 -->
00:02:06,407
So that allowed me to work across multiple
business lines and wear different hats,
40
00:02:06,537 -->
00:02:11,157
which involved building large scale
platforms from scratch, modernizing legacy
41
00:02:11,157 -->
00:02:13,037
platform onto these platforms, right?
42
00:02:13,037 -->
00:02:16,857
So thinking about opportunities
for new product development,
43
00:02:16,857 -->
00:02:19,807
new ways of accessibility that
comes from the modernization.
44
00:02:20,327 -->
00:02:24,815
So, the point being, while I was building
my engineering skills and building teams
45
00:02:24,815 -->
00:02:30,688
and I was, building departments, I was
also keeping track of customer problems,
46
00:02:30,708 -->
00:02:32,398
driving the customer listening sessions.
47
00:02:32,448 -->
00:02:35,478
And I was also lucky because of that
I got to travel internationally.
48
00:02:36,048 -->
00:02:39,728
But one thing that really taught me
that as an early career engineer there
49
00:02:40,118 -->
00:02:43,918
was the importance of shifting left on
the scale of product life cycle, right?
50
00:02:44,448 -->
00:02:46,488
And really develop that product sense.
51
00:02:46,968 -->
00:02:51,298
I think product sense is such a crucial
skill in all high performing engineering
52
00:02:51,338 -->
00:02:55,308
teams that I've worked with that it
definitely needs to be underscored.
53
00:02:55,692 -->
00:02:58,932
So coming to the experience, the
platforms we built there, especially
54
00:02:58,932 -->
00:03:02,627
towards the later part of my career
in SunGard, were with derivatives
55
00:03:02,627 -->
00:03:08,007
processing, but the focus was on high
throughput, low latency types of systems.
56
00:03:08,357 -->
00:03:11,667
You're now thinking about in memory
database, how to avoid context switching
57
00:03:11,667 -->
00:03:15,347
and from processes to processes to
get the best performance you can.
58
00:03:15,847 -->
00:03:19,177
While we were doing that, I got to
build these platforms from scratch for
59
00:03:19,177 -->
00:03:21,017
data processing and data connectivity.
60
00:03:21,597 -->
00:03:24,577
And then eventually also build an
API store for our customers to make
61
00:03:24,577 -->
00:03:26,667
it all available for integration.
62
00:03:27,367 -->
00:03:30,737
So that was a great experience, and
now I'm thinking about those concepts
63
00:03:30,737 -->
00:03:31,917
similarly in healthcare, right?
64
00:03:31,967 -->
00:03:35,937
So security, compliance, handling
of sensitive data were all critical
65
00:03:36,057 -->
00:03:40,037
in those environments, and those
principles translate directly
66
00:03:40,037 -->
00:03:41,117
to healthcare as well, right?
67
00:03:41,507 -->
00:03:45,867
When we think about financial privacy
and integrity, it's crucial we think
68
00:03:45,867 -->
00:03:51,007
about those things even more when we
think about patient records, right?
69
00:03:51,507 -->
00:03:55,707
I also think about Fintech, and even
other industries that I've looked
70
00:03:55,707 -->
00:04:00,437
at in general, the place of pace
of innovation feels faster because
71
00:04:00,447 -->
00:04:02,347
stakes are a little bit different.
72
00:04:02,997 -->
00:04:07,687
Even with regulation, there are no
patient lives at stake there, right?
73
00:04:07,687 -->
00:04:10,127
So pace definitely feels
and looks different.
74
00:04:11,007 -->
00:04:15,247
I do feel that API-driven ecosystem
and modern infrastructure were
75
00:04:15,247 -->
00:04:18,097
far more advanced in Fintech
before I got to the AMA.
76
00:04:18,647 -->
00:04:23,507
And I kind of think that kind of led
me to think about how can I think
77
00:04:23,517 -->
00:04:26,367
about scalability and flexibility
into some of the problems I'll be
78
00:04:26,367 -->
00:04:27,977
solving here into healthcare, right?
79
00:04:28,357 -->
00:04:31,207
Think about iterative value
creation think about unifying
80
00:04:31,207 -->
00:04:33,077
data assets and accessibility.
81
00:04:33,467 -->
00:04:36,907
Those things translate really
well into healthcare too, right?
82
00:04:37,187 -->
00:04:38,927
So that's how I think about those things.
83
00:04:40,541 -->
00:04:42,881
Gordon Wong: You underscored
a familiar distinction for me.
84
00:04:43,261 -->
00:04:45,511
Like, frequently as engineers,
we talk about things like
85
00:04:46,161 -->
00:04:49,721
velocity or total volume, real
time analytics and so on, right?
86
00:04:49,981 -->
00:04:54,371
And so, as engineers, we might think
that fintech and healthcare are
87
00:04:54,371 -->
00:04:57,581
different, but what I just heard
you say is that when you bubble out
88
00:04:57,581 -->
00:05:01,211
to the problem, the larger problem,
they actually have a lot in common.
89
00:05:01,261 -->
00:05:01,751
Is that right?
90
00:05:02,461 -->
00:05:03,401
Sandeep Dhamale: Yep, that is correct.
91
00:05:03,491 -->
00:05:07,251
And that's one thing, a cool thing about
being an engineer, you're able to look
92
00:05:07,251 -->
00:05:12,631
at the commonalities across the problem
spaces and bring on thinking that really
93
00:05:12,911 -->
00:05:14,341
makes you see the art of possible.
94
00:05:14,691 -->
00:05:17,391
Because you, if you've solved a
problem a certain way, why can't
95
00:05:17,391 -->
00:05:18,521
you solve it for this industry?
96
00:05:18,521 -->
00:05:21,115
Those are the kind of things
that engineers always love.
97
00:05:21,115 -->
00:05:25,071
And I, I have seen those commonalities
first hand for sure, especially when
98
00:05:25,071 -->
00:05:29,151
it comes to, uh, solving for data
unifying all of those data assets to
99
00:05:29,151 -->
00:05:30,651
leverage the value from that data.
100
00:05:32,711 -->
00:05:35,831
Gordon Wong: Multiple years in the
fintech world, built some really cool
101
00:05:35,831 -->
00:05:37,921
solutions, now you're in healthcare.
102
00:05:38,351 -->
00:05:41,681
Healthcare data comes with unique
challenges, such as silos and compliance,
103
00:05:42,101 -->
00:05:45,071
legacy systems, heterogeneous systems.
104
00:05:45,121 -->
00:05:47,481
What are the foundational
hurdles you've encountered, and
105
00:05:47,481 -->
00:05:48,361
how have you addressed them?
106
00:05:49,621 -->
00:05:50,131
Sandeep Dhamale: Right.
107
00:05:50,391 -->
00:05:55,381
So foundational hurdles, like some of
the things that you've mentioned is
108
00:05:55,701 -->
00:06:01,401
healthcare is in a state what it is
because of various different reasons.
109
00:06:01,461 -->
00:06:05,141
Of course compliance and speed
being one of the reasons, but
110
00:06:05,141 -->
00:06:09,031
because of that the problem spaces
of legacy infrastructure exist
111
00:06:09,341 -->
00:06:09,941
a lot more.
112
00:06:09,941 -->
00:06:14,501
So data fragmentation has creeped in a
lot in my opinion because there are so
113
00:06:14,501 -->
00:06:18,321
many different systems, they're siloed,
the data doesn't talk to each other.
114
00:06:18,641 -->
00:06:22,421
That has been one common theme that
I've noticed across healthcare.
115
00:06:22,751 -->
00:06:27,431
It's also a lot of build versus
buy challenges that come into
116
00:06:27,431 -->
00:06:31,391
places and decision making that
is happening in a siloed manner.
117
00:06:31,401 -->
00:06:36,821
So there's duplication of platforms data
being replicated across multiple places.
118
00:06:36,831 -->
00:06:41,011
So there's that redundancy that's
happening which actually some of the
119
00:06:41,021 -->
00:06:46,321
recent regulation like HIPAA and GDPR
are only putting a more spotlight on
120
00:06:46,851 -->
00:06:51,191
like if you look at where is your right
to forget or if you go to look for those
121
00:06:51,201 -->
00:06:55,661
records, it's not just in one place,
but it's in like 26 different places
122
00:06:55,661 -->
00:06:57,361
across the enterprise that you learn.
123
00:06:57,391 -->
00:07:02,191
And then, then Joel on some days
emailing spreadsheets of the same
124
00:07:02,191 -->
00:07:05,671
data to somewhere and that only
exacerbates the problem, right?
125
00:07:05,671 -->
00:07:09,091
So I think the hurdle was the
fragmentation and trying to think
126
00:07:09,091 -->
00:07:14,021
about how can we organize this data
in a centralized place so that we
127
00:07:14,031 -->
00:07:18,021
have a better control and better data,
unified data management in place.
128
00:07:18,031 -->
00:07:21,281
So that was one of the first
ones that I think about.
129
00:07:21,681 -->
00:07:25,671
I really also think about governance
frameworks and compliance for designs
130
00:07:25,701 -->
00:07:27,711
is becoming very, very crucial.
131
00:07:28,081 -->
00:07:33,001
But while I emphasize those, I also
want to emphasize flexibility for
132
00:07:33,001 -->
00:07:38,231
growth and thinking about scalability
in mind are as important as well.
133
00:07:38,231 -->
00:07:42,161
So yeah, I, I think the hurdles
are really data fragmentation and
134
00:07:42,161 -->
00:07:45,251
interoperability and how do we
really get those all together.
135
00:07:47,221 -->
00:07:50,701
Gordon Wong: You know, at Newfire, one
of the things I do with our clients
136
00:07:50,711 -->
00:07:54,801
is I advise them to think about the
ROI of their data efforts, right?
137
00:07:55,061 -->
00:07:58,811
You have to consider what are you going
to get out of this and what are the costs.
138
00:07:59,111 -->
00:08:02,371
Now, in fintech, it's probably a little
bit easier to measure that, right, because
139
00:08:02,371 -->
00:08:03,871
you're looking at financial returns.
140
00:08:04,671 -->
00:08:07,481
Do you have any suggestions
on how to measure the ROI of
141
00:08:08,171 -->
00:08:09,631
data efforts in healthcare?
142
00:08:10,836 -->
00:08:11,336
Sandeep Dhamale: Sure.
143
00:08:11,386 -->
00:08:13,117
You can look at it two
different ways, right?
144
00:08:13,167 -->
00:08:17,137
Definitely start with your use case
in mind and what you're really what
145
00:08:17,157 -->
00:08:21,557
personas you're trying to serve, because
that's going to determine your ROI
146
00:08:21,707 -->
00:08:25,177
case, because in some cases it's a data
product that you're building, which is
147
00:08:25,187 -->
00:08:27,147
going to have a real financial incentive.
148
00:08:27,207 -->
00:08:28,397
And that's why you're building it.
149
00:08:28,747 -->
00:08:31,787
In some cases, you're really thinking
about administrative use cases.
150
00:08:31,787 -->
00:08:34,207
And you're really trying
to think about how.
151
00:08:34,387 -->
00:08:38,217
How this problem will make everybody's
life easier and get the type of
152
00:08:38,217 -->
00:08:40,197
compliance and processes that we need.
153
00:08:40,257 -->
00:08:45,687
And ROI in, when I was building
infrastructure, we've done it both ways.
154
00:08:45,847 -->
00:08:50,077
Uh, one is thinking about what is
our current total cost of ownership
155
00:08:50,077 -->
00:08:53,137
of these siloed data systems and
how much are we really spending.
156
00:08:53,767 -->
00:08:58,402
And two, thinking about when we get
to our target vision of a unified data
157
00:08:58,402 -->
00:09:02,172
infrastructure— what is the cost of
running the business going to look like,
158
00:09:02,432 -->
00:09:04,302
and how do we transition from A to B.
159
00:09:04,302 -->
00:09:06,482
And there are always
so many wins to be had.
160
00:09:06,492 -->
00:09:10,932
You're, you almost always certainly
decrease your cost and reduce your risk.
161
00:09:11,252 -->
00:09:14,122
So that's that's a win with
cloud based architectures.
162
00:09:14,122 -->
00:09:16,952
You're always guaranteed more
and more flexibility that kind of
163
00:09:17,422 -->
00:09:19,082
prepares you for the future evolution.
164
00:09:19,572 -->
00:09:19,952
And.
165
00:09:20,582 -->
00:09:24,592
And then you realize that while you're
building this infrastructure, if you
166
00:09:24,672 -->
00:09:29,762
attach a couple of data product or real
revenue generating use cases to back
167
00:09:29,762 -->
00:09:34,702
it up, you really bake in the cost of
maintaining this infrastructure that
168
00:09:34,702 -->
00:09:38,972
kind of gets offset by the, by those
revenue generating products that you've
169
00:09:39,022 -->
00:09:43,822
included in the first cut of the draft
that you're thinking about the vision.
170
00:09:44,382 -->
00:09:48,482
And, and it has now unlocked value
because you can build the product 3, 4,
171
00:09:48,552 -->
00:09:53,527
5, 6 fairly quickly with reusing all of
the infrastructure that you've really
172
00:09:53,527 -->
00:09:55,207
built in for first couple of things.
173
00:09:55,547 -->
00:09:57,377
And then the ROI really scales from there.
174
00:09:57,617 -->
00:10:01,767
So definitely when you're thinking about
it, have those principles in mind is what
175
00:10:01,767 -->
00:10:06,637
I is what has benefited me and I think
should benefit our listeners as well.
176
00:10:07,224 -->
00:10:09,104
Gordon Wong: We use a lot of
metaphors in our industry, right?
177
00:10:09,104 -->
00:10:13,744
We talk about DAGs, directed graphs,
we talk about pipelines, we talk
178
00:10:13,754 -->
00:10:17,704
about manufacturing moving data into
knowledge, but more and more I'm
179
00:10:17,704 -->
00:10:21,584
thinking that the correct metaphor is
that this is actually a marketplace.
180
00:10:22,514 -->
00:10:25,744
We have sources on one side and we
have consumers on the other side and
181
00:10:25,754 -->
00:10:31,014
we're trying to enable transactions
and any good marketplace should take
182
00:10:31,014 -->
00:10:32,744
as little of a cut as possible, right?
183
00:10:33,069 -->
00:10:33,619
Sandeep Dhamale: Yes.
184
00:10:33,974 -->
00:10:34,284
Gordon Wong: Right?
185
00:10:34,464 -->
00:10:36,594
So does that, does that
metaphor hold in healthcare?
186
00:10:36,594 -->
00:10:37,844
Do you think, do you find that valuable?
187
00:10:38,969 -->
00:10:39,909
Sandeep Dhamale: I do actually.
188
00:10:39,969 -->
00:10:45,299
The more and more I think about how to
bring value to the data, it's really to
189
00:10:45,329 -->
00:10:49,984
try to think about how you're really going
to cater this data and what packaging
190
00:10:50,274 -->
00:10:54,214
and what formats and how little or
how big of that package needs to be.
191
00:10:54,214 -->
00:10:56,094
So you really have to
think about the personas.
192
00:10:56,524 -->
00:10:59,864
One of the other thing that's
where data marketplaces are
193
00:10:59,874 -->
00:11:02,764
becoming so useful in my mind.
194
00:11:03,074 -->
00:11:05,724
And it also shifts to the
healthcare because I think the
195
00:11:05,734 -->
00:11:07,014
problem is, not very different.
196
00:11:07,014 -->
00:11:08,444
You're thinking about interoperability.
197
00:11:08,444 -->
00:11:13,184
You're thinking about making it more
accessible and accessible is the is key
198
00:11:13,224 -->
00:11:16,954
when you think about marketplace like
concepts is like you're coming to a place.
199
00:11:17,164 -->
00:11:19,524
I always think of it like
a grocery store, right?
200
00:11:19,824 -->
00:11:23,754
You go to a grocery store aisle you,
you can pick up a you, you know exactly
201
00:11:23,754 -->
00:11:24,944
what you're doing in the grocery store.
202
00:11:24,944 -->
00:11:28,674
You know where to find, Your vegetables,
where you want to find packaged
203
00:11:28,674 -->
00:11:30,574
foods or where you want to find milk.
204
00:11:30,574 -->
00:11:33,874
So there, it's all well organized,
where the dairy section is and all that.
205
00:11:34,214 -->
00:11:37,144
You can go to that section, you pick
up a product of the shelf, you can
206
00:11:37,144 -->
00:11:41,024
read all the labels, you know what
you're getting and how it was sourced,
207
00:11:41,024 -->
00:11:44,514
etc, etc, what the ingredients are,
or the nutrition information, right?
208
00:11:44,564 -->
00:11:47,874
It's readily available for you to make
your decisions right there and there.
209
00:11:47,894 -->
00:11:52,014
So you don't feel very confused as to
what you're going to walk out with.
210
00:11:52,034 -->
00:11:55,314
You're going to walk out with a
whole milk or a 2 percent milk,
211
00:11:55,514 -->
00:11:56,344
those are the kind of things.
212
00:11:56,624 -->
00:11:58,784
So I also think about data
marketplaces the same way.
213
00:11:58,784 -->
00:12:02,549
If you have your data products
very well listed and what you're
214
00:12:02,549 -->
00:12:04,309
getting is clear for the consumer
215
00:12:04,339 -->
00:12:05,119
that's a win.
216
00:12:05,129 -->
00:12:09,049
That's how you should think about
packaging your data products in a way that
217
00:12:09,089 -->
00:12:13,519
a consumer comes to a data marketplace
that they're getting what they're needing.
218
00:12:14,229 -->
00:12:15,639
And that's one metaphor.
219
00:12:15,699 -->
00:12:19,429
And then the second metaphor also
that really resonated with me once
220
00:12:19,429 -->
00:12:21,669
was especially thinking about privacy.
221
00:12:22,029 -->
00:12:25,939
You take that grocery store and add
a pharmacy section to it, and now
222
00:12:25,939 -->
00:12:29,264
you need to have a prescription to
get a particular product, right?
223
00:12:29,734 -->
00:12:32,334
You're walking up to a pharmacy and
you're telling, I have a prescription.
224
00:12:32,334 -->
00:12:36,494
So that's to me most, some of the data
assets in that marketplace could be
225
00:12:36,494 -->
00:12:38,764
going that way because they're protected.
226
00:12:38,774 -->
00:12:41,494
They are important and
with sensitive information.
227
00:12:41,494 -->
00:12:43,784
So personas need to be going that way.
228
00:12:43,784 -->
00:12:44,874
So I agree with you.
229
00:12:45,214 -->
00:12:49,064
I think data products and the
marketplace's vision translates well,
230
00:12:49,154 -->
00:12:53,339
even into the healthcare ecosystem
because we're trying to really bring
231
00:12:53,339 -->
00:12:56,689
the same type of consumer experience to
healthcare that we've seen elsewhere.
232
00:12:58,119 -->
00:12:59,989
Gordon Wong: This is turning
into an economics podcast.
233
00:13:00,269 -->
00:13:02,940
Some of the concepts you brought up
earlier were definitely, you talked about,
234
00:13:03,399 -->
00:13:05,029
basically fixed and variable costs, right?
235
00:13:05,029 -->
00:13:10,189
And if costs are lower, you can cover, if
you have a baseline products that cover
236
00:13:10,579 -->
00:13:13,679
most of the costs of the platform, you
can afford to bring additional products.
237
00:13:13,829 -->
00:13:17,649
So that seems pretty key to scalability.
238
00:13:18,049 -->
00:13:18,299
All right.
239
00:13:18,359 -->
00:13:20,429
So I want to hear a little bit
more about some of the things
240
00:13:20,429 -->
00:13:21,659
you specifically built, right?
241
00:13:21,709 -->
00:13:24,199
Particularly, I to hear about
the AMA Intelligent Platform.
242
00:13:24,689 -->
00:13:25,609
Can you tell me about that?
243
00:13:26,888 -->
00:13:29,868
Sandeep Dhamale: it's going to be easy to
explain AMA Intelligent Platform because
244
00:13:29,868 -->
00:13:31,688
of some of the constructs we talked about.
245
00:13:32,028 -->
00:13:36,618
When I started at AMA in 2019 and
when we were looking at our future
246
00:13:36,618 -->
00:13:41,418
of data products, one of the ways we
were thinking about how can we bake
247
00:13:41,428 -->
00:13:46,133
in flexibility into an architecture
that can allow us to scale for future.
248
00:13:46,133 -->
00:13:49,183
So some of the exact same principles
that I described previously
249
00:13:49,513 -->
00:13:50,993
were were in consideration.
250
00:13:50,993 -->
00:13:55,593
We wanted to build a technical reusable
infrastructure on, cloud basically
251
00:13:55,593 -->
00:13:59,713
so that we can scale the way we want
to and add, keep adding capabilities.
252
00:13:59,973 -->
00:14:04,593
So think of it as, as more like a modern
data platform initiative where you're
253
00:14:04,633 -->
00:14:08,173
thinking about what is your unified
data strategy is going to be, where
254
00:14:08,173 -->
00:14:11,513
you're going to land the data, what kind
of tools you're going to bring in to
255
00:14:11,793 -->
00:14:16,493
process the data and create your gold
slash product repository, if you will
256
00:14:16,843 -->
00:14:18,563
what that infrastructure looks like.
257
00:14:18,888 -->
00:14:23,668
And then build capabilities on top of
it of creating different API endpoints,
258
00:14:23,698 -->
00:14:28,988
creating different type of integration
technologies for external facing whether
259
00:14:28,988 -->
00:14:33,538
they're external facing product consumers
or internal users even, but thinking
260
00:14:33,538 -->
00:14:35,518
about data accessibility from the get go.
261
00:14:36,333 -->
00:14:38,063
So it was a culmination of all of this.
262
00:14:38,063 -->
00:14:42,343
So it's baked with technical reusable
infrastructure in the back end with
263
00:14:42,343 -->
00:14:45,573
a digital front door on the front
end that allows our customers or
264
00:14:45,583 -->
00:14:49,853
even internal audiences to interact
with a developer portal built in.
265
00:14:49,873 -->
00:14:52,853
So you can you can look
at our API stack directly.
266
00:14:52,853 -->
00:14:57,573
So sort of, gives you a marketplace,
like a construct, but makes makes it easy
267
00:14:57,573 -->
00:15:01,513
for our customers to really interact with
our data and make it more accessible.
268
00:15:01,533 -->
00:15:04,153
That's, that, that was the vision
that we've been going in with,
269
00:15:04,343 -->
00:15:09,703
as we do more with this data, it
also creates a framework so for
270
00:15:09,733 -->
00:15:13,673
customers to predict what, how it's
going to look like and feel like.
271
00:15:13,993 -->
00:15:18,013
It's bringing some of the
standardization to access patterns.
272
00:15:18,013 -->
00:15:22,483
So that way, when we launch a new data
product, it's just going to be a walk in
273
00:15:22,483 -->
00:15:25,903
the park because they've done it already
once and it's going to look very similar.
274
00:15:25,903 -->
00:15:29,363
And the, all of the constructs that
they did with the first integration
275
00:15:29,553 -->
00:15:30,563
are going to just stand in there.
276
00:15:30,563 -->
00:15:30,803
Right.
277
00:15:30,803 -->
00:15:33,543
So that's how we think about
AMA Intelligent Platform.
278
00:15:33,893 -->
00:15:37,743
So for what it's worth, for a lot
of our newer data initiatives,
279
00:15:37,793 -->
00:15:40,973
they're all being facilitated by
this platform infrastructure now.
280
00:15:41,303 -->
00:15:44,213
And that has given us
flexibility for thinking about
281
00:15:44,213 -->
00:15:45,363
our data assets in the future.
282
00:15:46,323 -->
00:15:46,833
Gordon Wong: Incredible.
283
00:15:47,563 -->
00:15:50,828
So when you first, when the platform
was first conceived What were
284
00:15:50,828 -->
00:15:53,748
some of the real world problems
that you were hoping to solve?
285
00:15:53,798 -->
00:15:55,308
What, what kind of pain
were people facing?
286
00:15:56,493 -->
00:15:58,733
Sandeep Dhamale: I will categorize
into two sections, if you will.
287
00:15:58,733 -->
00:16:03,998
Uh, One is pains of how can we get our
infrastructure ready for innovation?
288
00:16:04,358 -->
00:16:06,688
Because we've had a lot of
legacy systems in place.
289
00:16:07,363 -->
00:16:11,153
And then the second is, of course,
the real customer voice, right?
290
00:16:11,153 -->
00:16:14,563
Like we've been hearing you,
you ship in certain formats.
291
00:16:14,563 -->
00:16:18,423
In a lot of cases, there were flat
files, you, we don't know how to
292
00:16:18,453 -->
00:16:23,223
track change history of these data
assets or in some cases, like it's not
293
00:16:23,223 -->
00:16:25,243
intuitive, the documentation around it.
294
00:16:25,583 -->
00:16:29,653
And it takes me a long time to
understand your data once it's delivered.
295
00:16:29,653 -->
00:16:31,963
So those were some of the key challenges.
296
00:16:31,963 -->
00:16:36,153
When we were thinking about architecture,
of course, we wanted to make sure we
297
00:16:36,153 -->
00:16:42,333
take a simpler approach of not trying
to replace everything at once, but
298
00:16:42,343 -->
00:16:46,243
think about how can we bring a data
platform that can be fed with some
299
00:16:46,243 -->
00:16:52,333
of the existing data assets that we
have to create an infrastructure to
300
00:16:52,343 -->
00:16:53,743
think about how to solve problems.
301
00:16:53,823 -->
00:16:56,763
And then from a customer listening
perspective, we were applying lenses
302
00:16:56,783 -->
00:17:01,093
as to, OK, you need more metadata
along with our existing data assets.
303
00:17:01,103 -->
00:17:03,013
So how can we generate
that metadata for you?
304
00:17:03,583 -->
00:17:07,013
OK, you need modern delivery mechanisms.
305
00:17:07,013 -->
00:17:10,823
In some cases, that just meant making
sure we have API delivery formats.
306
00:17:11,623 -->
00:17:13,213
Some were really low hanging fruit.
307
00:17:13,213 -->
00:17:16,573
Some were really thinking
about where this is going to go
308
00:17:16,583 -->
00:17:19,763
from , from our perspective, as
well as our customers perspective.
309
00:17:19,763 -->
00:17:23,593
One of the good things is
people really rely on AMS data.
310
00:17:23,633 -->
00:17:26,653
It's been powering
healthcare ecosystem for
311
00:17:29,233 -->
00:17:33,113
decades now, and we just want to make sure
we are good stewards of these data assets
312
00:17:33,463 -->
00:17:40,073
and continue to give it to our healthcare
partners and customers in the right
313
00:17:40,623 -->
00:17:45,833
time, right, format and, and make sure
that we continue to solve those challenges
314
00:17:45,833 -->
00:17:48,143
that we've solved in previous era as well.
315
00:17:49,613 -->
00:17:52,293
Gordon Wong: Now, I'm sure that every
project planned along the way was no
316
00:17:52,293 -->
00:17:53,783
problem, and you hit every deadline.
317
00:17:53,793 -->
00:17:55,393
But were there any challenges?
318
00:17:56,703 -->
00:18:00,103
Sandeep Dhamale: Of course, there's
no software project that ever
319
00:18:00,153 -->
00:18:02,023
throws a curveball at you, right?
320
00:18:02,033 -->
00:18:03,363
I think there were, right?
321
00:18:03,423 -->
00:18:08,053
But I think key to success in any
of these initiatives is making
322
00:18:08,053 -->
00:18:09,913
sure you have an end in mind.
323
00:18:10,113 -->
00:18:14,663
I always say be stubborn on your vision,
but be flexible on details and take,
324
00:18:14,813 -->
00:18:18,673
think about how your iterations could
look different based on your priorities
325
00:18:19,038 -->
00:18:23,138
and what personas so your
scope can be changed as you're
326
00:18:23,138 -->
00:18:24,478
learning about the marketplace.
327
00:18:24,478 -->
00:18:27,428
So think about what your milestones
and their definitions can be
328
00:18:27,748 -->
00:18:29,928
can be and still add a value.
329
00:18:29,948 -->
00:18:34,378
So definitely go incremental smaller
iterations and think about how can
330
00:18:34,378 -->
00:18:36,308
you ship a value every iteration.
331
00:18:36,743 -->
00:18:40,473
And then for your bigger
milestones, be flexible as you're
332
00:18:40,473 -->
00:18:41,893
learning from the marketplace.
333
00:18:42,203 -->
00:18:47,843
I can tell you, for example, there were
certain use cases that were not, that
334
00:18:47,853 -->
00:18:50,843
we thought were high value early on.
335
00:18:51,173 -->
00:18:55,203
And then to, to some of the market
research and even internal hurdles
336
00:18:55,503 -->
00:18:58,303
that we thought, okay, it's going
to take us longer to launch this.
337
00:18:58,648 -->
00:19:02,018
But we were able to pivot that
into a different type of package
338
00:19:02,278 -->
00:19:05,408
and make sure make sure we still
deliver value to our customers.
339
00:19:05,738 -->
00:19:09,118
But also our infrastructure
initiative doesn't get stalled
340
00:19:09,418 -->
00:19:12,328
because you always need to make
sure your customers are happy.
341
00:19:12,358 -->
00:19:16,488
And that kind of gives us the
impetus and funding, in fact, also
342
00:19:16,488 -->
00:19:20,118
to make sure that you can self
sustain and self fund the platform
343
00:19:20,118 -->
00:19:21,488
initiative that you've started here.
344
00:19:23,398 -->
00:19:25,568
Gordon Wong: Would you be willing
to share some of your wins with us?
345
00:19:25,718 -->
00:19:26,648
Feel free to brag.
346
00:19:28,593 -->
00:19:33,223
Sandeep Dhamale: I really think the
biggest one is is organizational one
347
00:19:33,513 -->
00:19:37,843
is taking everybody along on this,
and I think some of the builders
348
00:19:37,843 -->
00:19:41,913
will relate to it is it's one
thing to make, know what is right.
349
00:19:42,213 -->
00:19:46,363
And then you having a vision and
being able to also evangelize that
350
00:19:46,363 -->
00:19:48,513
across it takes a village, right?
351
00:19:48,523 -->
00:19:52,473
So you have to really be able
to talk to all the departments
352
00:19:52,583 -->
00:19:53,773
and all the stakeholders.
353
00:19:54,193 -->
00:19:59,093
And making sure that this this resonates
with everyone, and this is the most
354
00:19:59,093 -->
00:20:02,823
important problem that you're trying
to, so the framing of this initiative
355
00:20:03,193 -->
00:20:07,633
it took us a little bit I'm going to
say a lot of attritions to get it right,
356
00:20:07,673 -->
00:20:11,953
but I think I feel very good about it
right now that we are on a right path.
357
00:20:12,043 -->
00:20:13,543
Uh, so that's one.
358
00:20:13,983 -->
00:20:19,154
From an infrastructure and technical
perspective, I think getting our
359
00:20:19,234 -->
00:20:24,094
developer program stood up was one
of the highlights of this initiative.
360
00:20:24,454 -->
00:20:26,084
So I'll give you a back story, right?
361
00:20:26,084 -->
00:20:28,284
Like, we were building this
infrastructure for all the
362
00:20:28,284 -->
00:20:29,504
reasons that I mentioned to you.
363
00:20:29,874 -->
00:20:36,544
And then we had an initiative where we
wanted to make our CPT asset available
364
00:20:36,544 -->
00:20:43,124
to early stage builders with an open
license kind of a thing, where they can
365
00:20:43,124 -->
00:20:47,444
get access to our CPT content when they're
building their use cases and they've
366
00:20:47,444 -->
00:20:50,624
not figured out their go to market, for
example, and they want to have access.
367
00:20:51,074 -->
00:20:54,394
It was a perfect marriage because
we had an infrastructure in place.
368
00:20:54,464 -->
00:20:58,384
All we had to do was to think about
what a CPT developer program could look
369
00:20:58,384 -->
00:21:02,804
like, invite these builders onto the
AMA intelligent platform, and reuse the
370
00:21:02,804 -->
00:21:04,624
same infrastructure that we have, right?
371
00:21:04,914 -->
00:21:09,354
We have a developer portal, we have
an API store, and that we had built
372
00:21:09,374 -->
00:21:12,724
for our and curated for our customers,
but we were able to repurpose
373
00:21:13,224 -->
00:21:16,194
and launch a CPT developer
program in matter of weeks.
374
00:21:16,224 -->
00:21:20,274
Like I, I think from conception
to reality, it was like four to
375
00:21:20,274 -->
00:21:24,484
six weeks of project and boom, it
was launched and it's one of the
376
00:21:24,484 -->
00:21:30,169
programs that has given us a community
that actually participates with us.
377
00:21:30,289 -->
00:21:32,659
We were able to take
those connections forward,
378
00:21:32,659 -->
00:21:36,109
we interacted with our conferences,
we have quarterly Dev Chats.
379
00:21:36,109 -->
00:21:40,619
And they've been actually, uh, some of
the people who are working on cool use
380
00:21:40,639 -->
00:21:42,399
cases with our content and data products.
381
00:21:42,399 -->
00:21:46,799
So it's always useful to hear perspective
because industry is like you have matured.
382
00:21:47,649 -->
00:21:50,119
Participants that are using your
data assets and there are some
383
00:21:50,129 -->
00:21:52,559
new use cases you're seeing, like
how people are thinking about
384
00:21:52,559 -->
00:21:53,769
what they can do with your data.
385
00:21:54,119 -->
00:21:57,129
And having that voice baked
in has really helped us.
386
00:21:57,129 -->
00:22:01,559
So I think that was one of the
coolest achievement of AMA Intelligent
387
00:22:01,559 -->
00:22:04,989
Platform was to be able to launch a
CPD developer program, for example.
388
00:22:05,029 -->
00:22:05,229
Yeah.
389
00:22:05,279 -->
00:22:12,089
I think what's next is, is definitely
AI and generative AI on our minds.
390
00:22:12,509 -->
00:22:18,589
I also think it's, it was a good test
when AI wave came along is some of the
391
00:22:18,589 -->
00:22:23,259
investments that we'd started making in
this infrastructure stands true or not.
392
00:22:23,369 -->
00:22:24,689
And I think it has, right?
393
00:22:24,739 -->
00:22:31,034
If you think about, doing your AI or, more
importantly, generative AI now, right?
394
00:22:31,704 -->
00:22:36,094
You need to have the infrastructure that
backs it and you need to be ready with
395
00:22:36,404 -->
00:22:41,864
your data in a unified store and the right
formats and the right point of delivery.
396
00:22:42,874 -->
00:22:47,004
And I think this initiative has
started to push us in that direction.
397
00:22:47,434 -->
00:22:52,074
Some of the interesting things that
we're thinking about there has been
398
00:22:52,794 -->
00:22:58,979
more about unstructured content that
have been sitting in PDFs and different
399
00:22:58,979 -->
00:23:03,439
because a lot of our data assets were
also, one of our business, I'm going
400
00:23:03,439 -->
00:23:07,819
to say, was really a book business
or a human oriented business, which
401
00:23:07,819 -->
00:23:09,289
is now turning into a data business.
402
00:23:09,299 -->
00:23:10,009
So that's a shift.
403
00:23:10,339 -->
00:23:13,609
So we're taking some of those
assets and trying to trying to
404
00:23:13,699 -->
00:23:18,809
really create RAG architectures or
graph databases and vector stores.
405
00:23:19,149 -->
00:23:23,279
Those technologies, we were able to
onboard onto our data platform in
406
00:23:23,279 -->
00:23:26,709
fairly a quick amount of time because
we had started landing all of this
407
00:23:26,989 -->
00:23:28,509
into our data platform already.
408
00:23:28,799 -->
00:23:32,419
This was just, this is the good part
about modern data platform that I
409
00:23:32,419 -->
00:23:37,349
think is you can bring another piece of
technology into your stack fairly easily
410
00:23:37,679 -->
00:23:41,249
and it's able to, like you're thinking
about your architecture all the time
411
00:23:41,249 -->
00:23:43,629
is volume, velocity and variety, right?
412
00:23:43,629 -->
00:23:47,789
So this variety of, the new variety
of data and how do you really bring
413
00:23:47,789 -->
00:23:52,139
it to bring it to take advantage of
vector storage, for example, what was,
414
00:23:52,139 -->
00:23:54,629
is something that we're working on, right?
415
00:23:54,629 -->
00:23:55,679
Like you asked me what's next.
416
00:23:55,689 -->
00:23:57,929
So this is definitely on our minds.
417
00:23:58,279 -->
00:24:01,259
And there are some couple of use cases
that we can maybe talk about that
418
00:24:01,259 -->
00:24:03,139
we're thinking in that space as well.
419
00:24:03,189 -->
00:24:03,349
Yeah.
420
00:24:04,849 -->
00:24:06,289
Gordon Wong: Yeah, well,
I'd love to hear those.
421
00:24:07,269 -->
00:24:07,709
Sandeep Dhamale: Sure.
422
00:24:07,749 -->
00:24:12,229
I mean, yeah with AI, when a couple of
years ago, I think it's coming up on
423
00:24:12,229 -->
00:24:18,504
the second anniversary now when Chat
GPT launched we, I think early on, we
424
00:24:18,504 -->
00:24:22,014
pretty much jumped in with thinking
about what use cases and what are the
425
00:24:22,014 -->
00:24:24,534
framing that we want to put in place.
426
00:24:24,904 -->
00:24:28,334
I think we were bucketing our use
cases into a everyday AI and a
427
00:24:28,334 -->
00:24:30,364
transformative AI type of category.
428
00:24:30,884 -->
00:24:35,714
With everyday I mean, can AI help
us automate some of our tasks, make
429
00:24:35,714 -->
00:24:39,854
lives easy for our people give them
hours back in the day and solve some
430
00:24:39,854 -->
00:24:43,714
of the search problems even that
we had with these assets, right?
431
00:24:43,994 -->
00:24:49,004
So that's where we focused first is is
trying to train a model and create a
432
00:24:49,004 -->
00:24:53,624
RAG infrastructure on our book business
oriented data asset to see whether we
433
00:24:53,624 -->
00:24:58,084
can unlock some real value and create
an assistant for our internal staff
434
00:24:58,484 -->
00:25:03,584
to ask questions and see if they can
find right documents and references.
435
00:25:03,594 -->
00:25:08,264
So it's a combination of a search
problem and a a really easy chat based
436
00:25:08,274 -->
00:25:13,314
interface to think if this intelligence
layer can give you the answers
437
00:25:13,314 -->
00:25:16,954
right off the bat, because even this
internal team that I'm talking about
438
00:25:17,324 -->
00:25:21,084
gets questions from different teams
and even external people on guidance.
439
00:25:21,204 -->
00:25:24,184
So it is, the goal is to
accelerate their workflow.
440
00:25:24,214 -->
00:25:28,599
So that's a pilot that we're running
right now with the hope that this is the
441
00:25:28,599 -->
00:25:32,959
same infrastructure that we can actually
package intelligence into future and think
442
00:25:32,959 -->
00:25:38,009
about what kind of intelligence offerings
we can make into future products, right?
443
00:25:38,009 -->
00:25:41,309
So that's how that's an initiative
that we've been thinking about really.
444
00:25:41,419 -->
00:25:43,569
Gordon Wong: Yeah, so keep leading
into the marketplace, right?
445
00:25:43,569 -->
00:25:45,559
That's what it sounds like.
446
00:25:46,274 -->
00:25:46,959
Sandeep Dhamale: That is Correct
447
00:25:48,623 -->
00:25:50,513
Gordon Wong: One of the conversations
I have with our clients at
448
00:25:50,513 -->
00:25:53,693
Newfire with I say, you know,
encourage 'em to think about ROI.
449
00:25:53,783 -->
00:25:56,663
They get very tired of hearing
me say ROI over and over again.
450
00:25:56,663 -->
00:25:58,223
But that's, that's really
kind of everything.
451
00:25:58,253 -->
00:26:01,343
We have both the, the numerator,
the denominator, the value
452
00:26:01,343 -->
00:26:02,633
we generate and the costs.
453
00:26:02,663 -->
00:26:07,093
And I think personally I see a
tremendous amount of value in
454
00:26:07,093 -->
00:26:09,073
using AI to lower those costs.
455
00:26:09,163 -->
00:26:11,053
Lifting constraints, removing blockers.
456
00:26:11,473 -->
00:26:11,983
What are some.
457
00:26:12,308 -->
00:26:15,868
What are some areas of high friction
you see in delivering data products
458
00:26:15,868 -->
00:26:17,088
that you think AI can help with?
459
00:26:18,598 -->
00:26:21,898
Sandeep Dhamale: I think I've got a good
example for the question you're asking
460
00:26:21,948 -->
00:26:25,308
I mentioned we, we were a book business
and we're trying to get into a data
461
00:26:25,308 -->
00:26:27,168
business with that particular data asset.
462
00:26:27,798 -->
00:26:32,178
First challenge is making sure you
can segment the data the right way.
463
00:26:32,533 -->
00:26:32,763
Right?
464
00:26:32,763 -->
00:26:36,153
You have to break the content down
into a different format altogether.
465
00:26:37,213 -->
00:26:41,823
And then, you have to look for
common keywords, common tags,
466
00:26:41,873 -->
00:26:43,753
like create metadata on the fly.
467
00:26:44,103 -->
00:26:44,663
So,
468
00:26:44,884 -->
00:26:48,264
giving this to humans,
to curate a new data
469
00:26:48,474 -->
00:26:54,824
asset, reading through books, and
doing it like that, it's gonna be a
470
00:26:54,962 -->
00:26:56,522
It's going to take you forever to do it.
471
00:26:56,922 -->
00:27:01,662
I think what we're thinking about it is,
and we've done already some pilots with
472
00:27:01,662 -->
00:27:06,732
this is, can you use large language model
infrastructure to prompt it properly to
473
00:27:06,732 -->
00:27:12,642
create segments out of a big book pdf
and see what kind of output it generates.
474
00:27:13,032 -->
00:27:16,982
Then take those sections and create
like a preview summary that you can
475
00:27:16,982 -->
00:27:22,317
attach as a metadata tag into to that,
create keywords for that sections
476
00:27:22,317 -->
00:27:26,527
and generate like 10 keywords each
for that section and create that.
477
00:27:26,947 -->
00:27:32,297
And again, the goal is to accelerate
the stewardship of this new
478
00:27:32,297 -->
00:27:33,457
data asset that we're creating.
479
00:27:33,457 -->
00:27:37,257
So these are tools that we're
providing to our data teams or
480
00:27:37,257 -->
00:27:39,777
our content teams specifically.
481
00:27:40,242 -->
00:27:45,412
That who will be in charge of reviewing
this content and essentially stamping it.
482
00:27:45,412 -->
00:27:49,542
So these are tools that are going
to assist these data owners.
483
00:27:49,992 -->
00:27:54,572
To really look at what tags got generated,
how it's really creating sections of it.
484
00:27:54,572 -->
00:27:58,062
So it's not like you're taking and
fully automating it, but you're really
485
00:27:58,342 -->
00:28:02,232
making their lives easy and they are now
reviewers of this process rather than
486
00:28:02,262 -->
00:28:06,307
taking a book content and try to break
it down themselves one by one and reading
487
00:28:06,307 -->
00:28:07,717
through that's such a daunting task.
488
00:28:08,047 -->
00:28:12,647
I think AI is going to make that much more
simpler and thinking about curation and
489
00:28:12,647 -->
00:28:17,767
metadata creation process, that's one area
where you're gonna see your costs go down.
490
00:28:17,797 -->
00:28:22,217
And especially these initiatives
now are generating value, right?
491
00:28:22,217 -->
00:28:26,697
Like this was a value that
was sitting somewhere on a PDF
492
00:28:26,737 -->
00:28:29,027
document that was not getting
493
00:28:29,382 -->
00:28:34,212
uh, solved for now you have really
created a data offering out of it.
494
00:28:34,222 -->
00:28:37,642
So you have an opportunity to go
to market with a new offering.
495
00:28:38,092 -->
00:28:40,802
That's what we are seeing
with this kind of initiative.
496
00:28:40,852 -->
00:28:44,372
So whatever total addressable market
space you're looking at with that kind
497
00:28:44,372 -->
00:28:48,712
of content is definitely needs to be
factored into your ROI calculation.
498
00:28:48,712 -->
00:28:49,262
That's one.
499
00:28:49,612 -->
00:28:54,564
And then the second thing I'm going to
say is, because these data products are
500
00:28:54,574 -->
00:28:58,224
giving are going to the new markets,
you're going to get more feedback.
501
00:28:58,254 -->
00:29:04,264
For example for, in our case, a real
example, this was going into only
502
00:29:04,644 -->
00:29:08,994
into the hands of a certain section of
people who actually read this content.
503
00:29:09,334 -->
00:29:13,669
Now we're able to take these micro
Intelligence, if you will, and
504
00:29:13,669 -->
00:29:16,349
we're thinking about whether we
can inject into newer workflows.
505
00:29:16,649 -->
00:29:19,679
So that's like a newer market
you've you're thinking about now,
506
00:29:19,679 -->
00:29:22,559
not just the previous market that
you were working with, right?
507
00:29:22,929 -->
00:29:28,219
So I think you really have to be
able to re-imagine your domain and
508
00:29:28,219 -->
00:29:30,379
where it's heading in the age of AI.
509
00:29:30,619 -->
00:29:33,619
And then then think about
what possibilities exist and
510
00:29:33,629 -->
00:29:34,679
which bets you want to take.
511
00:29:34,699 -->
00:29:38,439
But yeah, there are going to be bets
that are going to be available to
512
00:29:38,439 -->
00:29:43,089
our, to like the listeners of this
podcast, because I think once you
513
00:29:43,089 -->
00:29:46,319
have data, you, the next layer you're
thinking about is intelligence, right?
514
00:29:46,319 -->
00:29:52,509
So I think the AI plays the role in
simplifying your data processing as well
515
00:29:52,509 -->
00:29:56,409
as thinking about new use cases where
your intelligence can be delivered.
516
00:29:56,529 -->
00:29:56,799
Yeah.
517
00:29:57,889 -->
00:30:00,789
Gordon Wong: As responsible stewards
of the platforms we build, I want
518
00:30:00,789 -->
00:30:03,759
to take this opportunity to remind
us of two expressions, right?
519
00:30:04,009 -->
00:30:05,399
One is knowledge is power.
520
00:30:06,059 -->
00:30:09,009
And then the other one is with great
power comes great responsibility.
521
00:30:09,289 -->
00:30:10,109
Sandeep Dhamale: Absolutely.
522
00:30:10,149 -->
00:30:15,919
Gordon Wong: So how do you see AI helping
us address the privacy concerns in
523
00:30:15,919 -->
00:30:17,239
healthcare and protecting our patients?
524
00:30:19,185 -->
00:30:20,735
Sandeep Dhamale: Great
question again, right?
525
00:30:20,795 -->
00:30:22,495
I I think two ways.
526
00:30:22,545 -->
00:30:26,815
One, and especially, I think we mentioned
this in somewhere in the podcast, right?
527
00:30:27,335 -->
00:30:30,195
When you're thinking about building
your data products, think with
528
00:30:30,195 -->
00:30:32,785
your end users in mind and the
personas you're building with.
529
00:30:33,315 -->
00:30:38,315
And they're going to be, once you
do your market research, you can
530
00:30:38,325 -->
00:30:42,775
actually work with large language
models to frame those personas right.
531
00:30:42,775 -->
00:30:45,355
And what kind of controls you
need in place, like defining
532
00:30:45,355 -->
00:30:48,755
them itself can be done with the
help of large language models.
533
00:30:48,755 -->
00:30:53,455
But I think, The daunting task of
automating the processes that you need to
534
00:30:53,465 -->
00:30:58,305
have in place for each of these personas,
like what are those processes going to
535
00:30:58,795 -->
00:31:03,955
be and then be able to create standard
operating procedures according around it.
536
00:31:04,305 -->
00:31:05,835
It's always a time problem, right?
537
00:31:05,875 -->
00:31:09,455
You don't have enough people to think
about think about these controls.
538
00:31:09,515 -->
00:31:12,905
You can automate a lot of these
controls with with help of AI.
539
00:31:12,945 -->
00:31:15,195
I think it's a problem.
540
00:31:15,555 -->
00:31:21,285
And a balanced space because AI is going
to make accessibility more prevalent.
541
00:31:21,875 -->
00:31:26,285
And then how can you also leverage
AI with security first mindset
542
00:31:26,635 -->
00:31:29,655
to think about automating the
processes and the governance
543
00:31:29,665 -->
00:31:31,645
framework around it is important.
544
00:31:31,675 -->
00:31:36,525
I think some of the things that you
talked about data marketplaces and how
545
00:31:36,545 -->
00:31:40,185
you can get your access streamlined
because of a data marketplace or that
546
00:31:40,185 -->
00:31:43,505
kind of an access pattern where people
are coming to the same place to get
547
00:31:43,505 -->
00:31:47,385
their data I think it still remains.
548
00:31:47,545 -->
00:31:52,155
And these are the kind of platform
thinking and processes will continue
549
00:31:52,155 -->
00:31:57,105
to make sure you have a governed
platform and not just an accessibility
550
00:31:57,155 -->
00:32:01,215
that has now created a newer type
of problem in security space.
551
00:32:02,645 -->
00:32:08,490
Gordon Wong: I like to find the
security people in whatever organization
552
00:32:08,490 -->
00:32:12,180
I'm in and ask them to teach me
how to be their best customer.
553
00:32:12,805 -->
00:32:15,375
Because otherwise, I'm just a walk,
I'm just a walking problem, right?
554
00:32:17,205 -->
00:32:17,515
Sandeep Dhamale: Yeah.
555
00:32:17,515 -->
00:32:22,505
Well, one thing, one thing that I think
especially in recent times, I've had a
556
00:32:22,585 -->
00:32:28,405
lot more appreciation is probably, and we
did this in our recent all hands within
557
00:32:28,565 -->
00:32:33,035
our engineering team is to start thinking
like everybody's a security engineer.
558
00:32:33,245 -->
00:32:34,415
That's part of your job.
559
00:32:34,735 -->
00:32:39,115
We've just tried to ingrain that
into our build teams mindset
560
00:32:39,515 -->
00:32:44,580
is our centralized security team is only
so much and they are going to evangelize
561
00:32:44,610 -->
00:32:49,270
the principles, but it's everybody's
responsibility and really thinking about
562
00:32:49,620 -->
00:32:52,210
that at the design time is very important.
563
00:32:52,700 -->
00:32:55,360
That also brings me to the
architecture point, right?
564
00:32:55,360 -->
00:32:59,230
Like evolution, evolvability
versus maintainability.
565
00:32:59,640 -->
00:33:01,500
And that's a, that's always a balance.
566
00:33:01,500 -->
00:33:04,390
You want to have an architecture
that scales and evolves.
567
00:33:04,805 -->
00:33:08,565
But it is also maintainable because
if it's not maintainable, your
568
00:33:08,575 -->
00:33:10,135
security is going to be a nightmare.
569
00:33:10,195 -->
00:33:15,205
And that's why we're trying to bring
a balance is it's your responsibility.
570
00:33:15,205 -->
00:33:16,785
It's everybody's responsibility.
571
00:33:16,785 -->
00:33:20,925
And we're trying to think about whether I,
I don't we've not done it, but we're now
572
00:33:20,925 -->
00:33:26,705
going to start putting like a line into
job descriptions also to start emphasizing
573
00:33:26,775 -->
00:33:30,955
this point more around security as
we think about further and further.
574
00:33:31,095 -->
00:33:31,235
Yeah.
575
00:33:32,955 -->
00:33:34,775
Gordon Wong: I'm going to take the
opportunity to jump on the soapbox.
576
00:33:34,825 -->
00:33:37,605
I'm going to try to frame
my statement as a question.
577
00:33:37,605 -->
00:33:43,885
But I personally have seen that it feels
like most organizations underinvest in the
578
00:33:43,885 -->
00:33:45,645
maintainability of their platforms, right?
579
00:33:45,645 -->
00:33:48,765
They focus on development costs,
but you don't really deliver value
580
00:33:48,765 -->
00:33:51,185
until the thing, until whatever
you're building is in production.
581
00:33:51,485 -->
00:33:52,525
Now, have you seen the same thing?
582
00:33:53,838 -->
00:33:58,488
Sandeep Dhamale: Yes, all the time we
see this around and it's also, I, like I
583
00:33:58,488 -->
00:34:03,978
was saying, it's a balance and sometimes
engineers have to be that voice of reason
584
00:34:04,378 -->
00:34:08,858
because I know everybody, including our
stakeholders are excited to see the value
585
00:34:09,178 -->
00:34:13,778
to market and time to market as the most
important thing and which is, which it is
586
00:34:13,778 -->
00:34:16,228
in my mind and you go through phases in.
587
00:34:16,963 -->
00:34:18,513
In my mind, you go through phases.
588
00:34:18,853 -->
00:34:22,593
If you're trying to build an MVP
that you want to head to the market
589
00:34:22,593 -->
00:34:27,533
quickly and see what it does, you
can think about you can think about
590
00:34:27,553 -->
00:34:29,763
what your maintainability looks like.
591
00:34:29,793 -->
00:34:33,993
But if you're making that decision, you
really have to make sure that you're
592
00:34:34,003 -->
00:34:38,923
baking it as a part of consideration
as to what what it takes to really get
593
00:34:38,933 -->
00:34:43,333
to the production grade product, if
you will and have that on your roadmap.
594
00:34:43,483 -->
00:34:44,813
You don't just discount it.
595
00:34:44,843 -->
00:34:47,123
You really make sure that it's out there.
596
00:34:47,633 -->
00:34:52,233
In my mind, You don't really put
maintainability completely out of
597
00:34:52,233 -->
00:34:56,573
the window from the beginning, you
actually bake it in, you do trade offs
598
00:34:56,783 -->
00:35:01,253
and try to work with those trade offs
based on your time to market, but have
599
00:35:01,253 -->
00:35:05,943
a roadmap, make sure it's out there
and everybody has seen it, that your
600
00:35:06,063 -->
00:35:09,753
MVP can hit with X parameters, but to
get to the production grade, you have
601
00:35:09,753 -->
00:35:15,193
to have X plus Y in place without that
you don't really go live is important.
602
00:35:15,293 -->
00:35:19,703
And it's also allowing them to
see why it's important, right.
603
00:35:19,733 -->
00:35:24,383
For ex I, I think that availability and
maintainability, I like to take an example
604
00:35:24,383 -->
00:35:26,963
all the time as a bicycle example, right?
605
00:35:27,013 -->
00:35:31,353
If I was asked to design a cycle
with a bike, with a requirement
606
00:35:31,353 -->
00:35:32,673
which is the most flexible,
607
00:35:33,068 -->
00:35:35,188
you can build like a monocycle, right?
608
00:35:35,218 -->
00:35:36,798
That one with a one tire.
609
00:35:37,328 -->
00:35:39,068
That one is pretty easy to maneuver.
610
00:35:39,068 -->
00:35:40,728
You can just do 360 on it.
611
00:35:41,128 -->
00:35:45,118
But it's not easy to ride
because balancing on it is hard.
612
00:35:45,558 -->
00:35:48,318
And then there's tricycle,
which you can build.
613
00:35:48,703 -->
00:35:51,243
Which has three wheels nobody's
going to fall off of it.
614
00:35:51,243 -->
00:35:53,773
It's very easy to balance,
but it doesn't go as fast.
615
00:35:53,773 -->
00:35:55,003
It's very hard to maneuver.
616
00:35:55,453 -->
00:35:58,113
And bicycle is like a
balance with two wheels.
617
00:35:58,113 -->
00:36:02,363
You can do balancing pretty well,
but you can also maneuver it.
618
00:36:02,703 -->
00:36:04,183
But if you look at it, it's a spectrum.
619
00:36:04,213 -->
00:36:07,213
Now you're making decision based
on what market you're operating
620
00:36:07,293 -->
00:36:08,853
and what situation you're in.
621
00:36:08,903 -->
00:36:12,513
If you're trying to really go fast
and you have a skilled guy who can
622
00:36:12,523 -->
00:36:18,243
actually do with one wheel, that's,
that can only sustain for a little bit.
623
00:36:18,333 -->
00:36:21,483
I think the bike is the most
prevalent architecture, right?
624
00:36:21,483 -->
00:36:25,133
You look for flexibility and it's
a balance because you don't want,
625
00:36:25,133 -->
00:36:30,933
also want a tricycle that doesn't go
as fast or can only cater to really
626
00:36:30,943 -->
00:36:35,343
specific niche of, in this case, small
kids who are really trying to learn
627
00:36:35,803 -->
00:36:37,563
to bike to get to the next level.
628
00:36:37,843 -->
00:36:39,903
And so, yeah, it's always
a balance, I think.
629
00:36:40,830 -->
00:36:41,120
Gordon Wong: I love that.
630
00:36:41,130 -->
00:36:41,680
I love that.
631
00:36:41,780 -->
00:36:42,180
I love that.
632
00:36:42,220 -->
00:36:45,940
And that's part of, I think that's all
part of the purposeful architecture of
633
00:36:45,960 -->
00:36:49,050
these products and these platforms is
that we have to understand the constraints
634
00:36:49,090 -->
00:36:52,600
and we need to make investments, measured
investments in the right places, right?
635
00:36:52,630 -->
00:36:54,230
You can't, you don't
get anything for free.
636
00:36:54,520 -->
00:36:54,900
So
637
00:36:55,875 -->
00:36:56,035
Sandeep Dhamale: That's
638
00:36:56,625 -->
00:36:56,895
true.
639
00:36:57,120 -->
00:36:58,510
Gordon Wong: What are
you excited about next?
640
00:36:58,570 -->
00:36:59,050
What's coming?
641
00:36:59,070 -->
00:37:01,620
What kind of emerging technologies
are coming that you can't
642
00:37:01,620 -->
00:37:02,410
wait to get your hands on?
643
00:37:04,069 -->
00:37:06,049
Sandeep Dhamale: I think I
kind of covered it with AI.
644
00:37:06,069 -->
00:37:11,689
So I'll tell you one thing that I think
it's going to happen is especially in
645
00:37:11,689 -->
00:37:15,019
the space that we're operating, all
of the data companies, and especially
646
00:37:15,039 -->
00:37:19,109
companies who have data, like data
that either they generate or the data
647
00:37:19,109 -->
00:37:23,214
that they have have been stewards of
for a long, long time, have a real
648
00:37:23,214 -->
00:37:27,204
opportunity to think about what additional
intelligence layer that they can build.
649
00:37:27,224 -->
00:37:28,634
So that's exciting in my mind.
650
00:37:28,634 -->
00:37:29,874
And we operate in that space.
651
00:37:29,874 -->
00:37:31,294
So very, very excited about that.
652
00:37:31,834 -->
00:37:35,164
The second thing is a lot of these
models that are out there have
653
00:37:35,164 -->
00:37:40,464
been trained on internet data
but not on enterprise data yet.
654
00:37:40,684 -->
00:37:45,214
So I think that's going to give rise to a
lot of domain specific models and domain
655
00:37:45,214 -->
00:37:48,414
specific use cases and healthcare is,
656
00:37:48,464 -->
00:37:52,524
being in healthcare, I think that gives me
a lot of hope to solve for problems that
657
00:37:52,534 -->
00:37:54,694
have been out of reach for a while now.
658
00:37:55,024 -->
00:37:57,674
Or just because of time
resource constraints.
659
00:37:57,674 -->
00:38:01,834
And when do you really get to it when you
have so many other things to solve, right?
660
00:38:01,864 -->
00:38:03,624
So I think that's exciting for me.
661
00:38:03,954 -->
00:38:07,764
And I think the third trend I'm
noticing is the private infrastructures.
662
00:38:08,114 -->
00:38:11,524
Because a lot of people who have
these data assets are kind of thinking
663
00:38:11,534 -->
00:38:16,634
about how can they not have to throw
this data out to these large language
664
00:38:16,644 -->
00:38:21,834
models where they're not sure of the
security and whether their data is not
665
00:38:21,854 -->
00:38:24,564
being trained for other purposes, etc.
666
00:38:24,564 -->
00:38:28,379
I think that's where private large
language model infrastructures are gonna
667
00:38:28,419 -->
00:38:32,519
play a key role in my mind and there are
vendors out there who are now supporting
668
00:38:32,519 -->
00:38:36,369
with with these initiatives but yeah,
and that was also one of the things
669
00:38:36,369 -->
00:38:40,819
that we did, just so you're aware, is
when when we build our infrastructure of
670
00:38:40,819 -->
00:38:43,749
the large language model that I talked
to you about a couple of use cases.
671
00:38:44,539 -->
00:38:49,059
With security first mindset we actually
because it was early days we got
672
00:38:49,069 -->
00:38:53,289
the open source model LLAMA 3 back
then and LLAMA 2 and LLAMA 3 now.
673
00:38:53,619 -->
00:38:57,129
And we have also looked at other
models to kind of bring on prem
674
00:38:57,129 -->
00:39:01,739
we've spun up a GPU infrastructure
ourselves and have been training and
675
00:39:01,739 -->
00:39:03,159
creating that infrastructure there.
676
00:39:03,609 -->
00:39:07,109
The with the whole goal, and I think
with newer frameworks and newer
677
00:39:07,109 -->
00:39:12,359
technologies, It's getting more easier
and easier to control your infrastructure
678
00:39:12,599 -->
00:39:14,199
off these large language models.
679
00:39:14,559 -->
00:39:19,449
It's still niche, but at least it's
possible that you are not now training
680
00:39:19,459 -->
00:39:22,459
your data and model elsewhere, right?
681
00:39:22,479 -->
00:39:24,979
You're, you can control
the, where that goes.
682
00:39:24,979 -->
00:39:27,459
It can stay within your network
or within your preferred
683
00:39:27,459 -->
00:39:28,969
partner network, if you will.
684
00:39:29,299 -->
00:39:31,379
And I think that's exciting to me.
685
00:39:31,379 -->
00:39:35,259
That kind of opens up a lot of use
cases because your security teams, if
686
00:39:35,259 -->
00:39:38,469
they're not as worried, at least then
you can think about leveraging large
687
00:39:38,469 -->
00:39:42,339
language infrastructure, large language
model infrastructure better for your use
688
00:39:42,339 -->
00:39:44,279
cases because we have so many of them.
689
00:39:44,319 -->
00:39:48,699
And I think that, that really was a,
you asked me about wins and brags.
690
00:39:48,699 -->
00:39:53,189
I think that was one of the wins and
brags that I definitely am proud of that
691
00:39:53,219 -->
00:39:59,364
we were able to tap into a private LLM so
early that kind of created possibilities
692
00:39:59,374 -->
00:40:04,224
to think about some of the data use
cases or the content to data use cases
693
00:40:04,504 -->
00:40:07,934
that we talked about and that's what
we'll be focusing on next year as well.
694
00:40:09,162 -->
00:40:11,402
Gordon Wong: I'm excited about
the same things, I find myself
695
00:40:11,472 -->
00:40:12,292
totally aligned with you.
696
00:40:12,532 -->
00:40:15,502
Sandeep I've asked you a lot of
questions in this 45 minutes or so.
697
00:40:15,502 -->
00:40:19,072
So I'm going to start a Newfire
podcast tradition right now.
698
00:40:19,172 -->
00:40:21,952
I think the guests should get a
chance to ask the host a question.
699
00:40:22,352 -->
00:40:23,462
Do you have any questions for me?
700
00:40:25,672 -->
00:40:26,382
Sandeep Dhamale: Sure, Gordon.
701
00:40:26,402 -->
00:40:28,602
First of all, it was
fun chatting with you.
702
00:40:28,602 -->
00:40:33,082
I think we talked a lot about data
and AI and you bring a very different
703
00:40:33,082 -->
00:40:36,122
perspective because you've been yourself
at the helm of so many different
704
00:40:36,122 -->
00:40:37,792
initiatives across your career.
705
00:40:38,272 -->
00:40:42,282
I think in your current role, you're
trying to help a lot of customers.
706
00:40:42,292 -->
00:40:46,693
So some of the things that I talked to
you about are, are those the same things
707
00:40:46,693 -->
00:40:48,733
that you're seeing across the industry?
708
00:40:48,733 -->
00:40:51,893
I was just curious, like your
perspective, because you have a
709
00:40:51,963 -->
00:40:55,823
vantage point like no other, and I'm
trying to get your perspective as
710
00:40:55,823 -->
00:41:00,603
well on, on the trends that you're
seeing across both data and AI space.
711
00:41:02,138 -->
00:41:03,398
Gordon Wong: Yeah, that's
a really good question.
712
00:41:03,418 -->
00:41:06,428
So one of the things I think I realized
when I think about it is that what's
713
00:41:06,438 -->
00:41:08,118
old is what's new again, right?
714
00:41:08,168 -->
00:41:14,038
So often what I'm finding is that we're
building very similar, not that I think
715
00:41:14,038 -->
00:41:18,198
it's exactly the same solution, but we're
solving the same problems now, at this
716
00:41:18,198 -->
00:41:20,188
point in my career, as I was 30 years ago.
717
00:41:21,158 -->
00:41:23,898
Because at the end of the day, what
we're trying to do is take data,
718
00:41:23,918 -->
00:41:26,378
some kind of measurement, some kind
of, and turn it into information and
719
00:41:26,378 -->
00:41:30,108
knowledge that allows better decisions,
better actions, and better outcomes.
720
00:41:30,368 -->
00:41:30,738
Right?
721
00:41:31,768 -->
00:41:35,445
And what we're seeing though is that as
the world's gotten more sophisticated
722
00:41:35,445 -->
00:41:38,125
and we have more technology and we
have more capabilities, we're just
723
00:41:38,135 -->
00:41:39,745
tackling more sophisticated problems.
724
00:41:40,375 -->
00:41:44,090
30 years ago, it might've been trying
to measure who my best customer is.
725
00:41:44,690 -->
00:41:49,980
But now we're trying to drive better
human health outcomes at scale, right?
726
00:41:50,110 -->
00:41:52,100
We're solving problems that
we've never solved before.
727
00:41:52,520 -->
00:41:55,570
And so that's what I think I'm
taking away from this and what I'm
728
00:41:55,570 -->
00:41:59,180
reminding my clients at Newfire,
and we talked to them is that, that
729
00:41:59,190 -->
00:42:01,900
yes, the technology is new, right?
730
00:42:01,900 -->
00:42:03,800
But the problems are old in some ways.
731
00:42:03,850 -->
00:42:05,110
It's just a different scale.
732
00:42:05,380 -->
00:42:09,270
So since we know how to solve these
problems, let's bring that thinking, that
733
00:42:09,270 -->
00:42:10,920
learning we've had in the past there.
734
00:42:11,100 -->
00:42:13,520
You know, ROI, we talked
about ROI a second ago, right?
735
00:42:13,660 -->
00:42:14,870
Numerator, denominator.
736
00:42:15,000 -->
00:42:15,370
Okay.
737
00:42:15,590 -->
00:42:18,100
So what are the costs that
we are incurring in trying
738
00:42:18,100 -->
00:42:19,180
to drive better outcomes?
739
00:42:19,540 -->
00:42:21,730
Okay, how do we burn those down, right?
740
00:42:22,010 -->
00:42:23,260
What's the value we're driving?
741
00:42:23,300 -->
00:42:24,430
Hey, what's holding back the value?
742
00:42:24,430 -->
00:42:26,730
Where's the friction that is
keeping us from delivering?
743
00:42:27,180 -->
00:42:31,910
If I tell my analysts, if you
deliver a presentation to your
744
00:42:31,910 -->
00:42:35,280
audience you give them advice on what
they should do in this situation.
745
00:42:35,600 -->
00:42:36,950
But they don't take that advice.
746
00:42:37,650 -->
00:42:39,300
You haven't really
generated any value yet.
747
00:42:39,890 -->
00:42:44,910
So part of your job as an analyst is
to be persuasive, is to communicate.
748
00:42:45,250 -->
00:42:48,850
And part of persuasion and communicating
is also proving that our advice or our
749
00:42:48,850 -->
00:42:52,060
data products are safe to use, right?
750
00:42:52,280 -->
00:42:55,840
You used the metaphor earlier
about milk and groceries, right?
751
00:42:56,110 -->
00:43:00,250
If I don't know this milk is safe to
drink, I'm not going to drink it, right?
752
00:43:00,430 -->
00:43:05,515
And if our customers don't know that
or our customers' stakeholders don't
753
00:43:05,515 -->
00:43:08,625
know that the data is safe to use,
they're not going to use it, right?
754
00:43:08,825 -->
00:43:11,525
So I encourage our clients to
make sure they are investing in
755
00:43:11,525 -->
00:43:15,075
that safety, that maintainability,
and make sure it's visible.
756
00:43:15,305 -->
00:43:17,525
And so when someone asks,
is your product safe to use?
757
00:43:17,525 -->
00:43:18,715
You say, absolutely, yes.
758
00:43:18,795 -->
00:43:20,045
And here's how I can prove it to you.
759
00:43:20,965 -->
00:43:21,255
Sandeep Dhamale: Yeah.
760
00:43:21,525 -->
00:43:25,075
Gordon Wong: We're seeing these same
things across all our company, all
761
00:43:25,075 -->
00:43:28,745
the companies and yes, a lot of use
cases in their details are new, a
762
00:43:28,745 -->
00:43:34,825
lot of chatbot stuff, a lot of using,
models to predict outcomes, right?
763
00:43:34,825 -->
00:43:37,715
And then we're making those outcomes
more deriving those predictions more
764
00:43:37,715 -->
00:43:42,215
quickly, it's greater accuracy further
into the future, but also tying all
765
00:43:42,215 -->
00:43:45,635
these assets together to get a more
holistic view , right, as opposed to
766
00:43:45,635 -->
00:43:49,285
just these little point views, if I
use a simplistic metaphor, if I'm going
767
00:43:49,285 -->
00:43:52,795
to the beach, I want to know how to
get there, what's the best clam shack,
768
00:43:52,825 -->
00:43:55,585
what the weather's going to be, what's
the tide, I want it all in one place.
769
00:43:55,585 -->
00:43:55,885
And
770
00:43:55,885 -->
00:43:56,055
it.
771
00:43:56,115 -->
00:43:56,995
now it's getting easier.
772
00:43:57,725 -->
00:43:58,315
Sandeep Dhamale: Yeah, no.
773
00:43:58,315 -->
00:43:58,695
Love it.
774
00:43:58,705 -->
00:43:59,005
Love it.
775
00:43:59,075 -->
00:44:00,985
And that totally resonates with me, right?
776
00:44:00,985 -->
00:44:05,415
Like, these are the same problems
that have existed, but the reach of
777
00:44:05,505 -->
00:44:10,770
our What the technology has enabled
is that it's now actually possible.
778
00:44:10,830 -->
00:44:17,910
I always think in early 2000s, it was
really hard to build a data platform
779
00:44:17,910 -->
00:44:21,390
that was, that can unify all of these
data assets with the advent of cloud
780
00:44:21,390 -->
00:44:25,260
infrastructure that kind of makes your
life much more easier to think about
781
00:44:25,280 -->
00:44:26,780
what it is going to look like, right?
782
00:44:26,780 -->
00:44:30,710
Like you have technologies at your
disposal, which can hold variety of
783
00:44:30,710 -->
00:44:35,280
data, which was harder back in 2000s
and with NoSQL and other technologies
784
00:44:35,600 -->
00:44:39,070
that have matured over the last
few years, it's become possible.
785
00:44:39,160 -->
00:44:40,640
Same story with Vector and AI.
786
00:44:40,660 -->
00:44:42,880
I think it's another tool
in your arsenal that's
787
00:44:43,135 -->
00:44:44,645
just gonna increase your reach.
788
00:44:44,695 -->
00:44:46,715
Really excited with those possibilities.
789
00:44:46,715 -->
00:44:50,815
And I totally love the images that
you use there, yeah, of the beach.
790
00:44:50,845 -->
00:44:50,985
Yeah.
791
00:44:50,985 -->
00:44:56,145
And let's with, with this, the beach,
beach hits home because I'm in Chicago
792
00:44:56,145 -->
00:44:58,915
right now and it's pretty, pretty gloomy.
793
00:44:58,915 -->
00:45:02,385
So I'm already thinking about
which beach I can be next.
794
00:45:04,095 -->
00:45:04,515
Yeah.
795
00:45:18,498 -->
00:45:18,655
Gordon Wong: IMe too, me too.
796
00:45:18,655 -->
00:45:18,842
Sandeep, it was fantastic talking to you.
797
00:45:18,842 -->
00:45:18,897
I really enjoyed myself.
798
00:45:18,897 -->
00:45:18,945
I feel like I actually learned something.
799
00:45:18,945 -->
00:45:18,968
I'm looking forward for us
to be working together again.
800
00:45:18,968 -->
00:45:18,975
And I hope you enjoyed yourself as well.
801
00:45:18,975 -->
00:45:19,518
Sandeep Dhamale: It was so much fun.
802
00:45:20,303 -->
00:45:22,943
I have always enjoyed
interacting with Newfire team.
803
00:45:23,163 -->
00:45:24,613
This is an amazing team.
804
00:45:25,643 -->
00:45:32,113
Everyone I've interacted with this at
this team has been super smart and I loved
805
00:45:32,163 -->
00:45:36,123
this conversation also about our data
infrastructure and how the AI initiatives
806
00:45:36,123 -->
00:45:39,813
and how to think about data and AI
infrastructure in this fast moving world.
807
00:45:40,149 -->
00:45:41,619
We should continue talking more.
808
00:45:41,629 -->
00:45:43,419
It's always fun to chat with you.
809
00:45:43,419 -->
00:45:44,579
Thank you for having me here.
810
00:45:45,254 -->
00:45:46,274
Gordon Wong: Of course, it's a pleasure.
811
00:45:46,524 -->
00:45:47,264
Thank you so much.
812
00:45:47,634 -->
00:45:48,163
Have a good day.
813
00:45:50,153 -->
00:45:50,803
Thanks for tuning in.
814
00:45:51,023 -->
00:45:53,753
Stay tuned for more episodes where
we continue to explore the toughest
815
00:45:53,753 -->
00:45:56,003
challenges and smartest solutions
in business and technology.
816
00:45:56,523 -->
00:45:57,323
Like and subscribe.
817
00:45:57,863 -->
00:46:00,493
Until next time, keep innovating
and solving the hard problems.
818
00:46:00,913 -->
00:46:01,823
This is Hard Problems.
819
00:46:02,013 -->
00:46:02,783
Smart Solutions.
820
00:46:02,953 -->
00:46:04,323
The Newfire Podcast.
