Latest / Identity at the Center / #317 - Sponsor Spotlight - Strivacity

Transcript

#317 - Sponsor Spotlight - Strivacity · Identity at the Center

0:00
A hallucination can kind of be thought of as a, as a statement
0:03
that's statistically probable, but factually incorrect.
0:07
I, I, I like that. I read that statement somewhere.
0:09
I was like, that's perfect way to describe it.
0:11
Statistically probable, factually incorrect, right.
0:13
And again, this is one of the reasons that people are building
0:17
grounding systems around AIS like like RAG, because it's a
0:20
safety mechanism, It's a security mechanism, like you
0:23
said, right? To make it more, make the answer
0:25
more reliable, make it more accurate.
0:28
It's certainly a concern for us as we were, as we were building
0:31
the product, building it into our product.
0:33
And it's also why you see disclaimers like on chatbot
0:37
systems. It's like, hey, LLMS can make
0:38
mistakes. Please, please verify your
0:40
answers because hallucinations are and will continue to be a
0:43
problem even with grounding systems around around, you know,
0:48
the, the implementation of these.
0:49
So yeah, I think it's, I think it's a great question.
0:52
Are hallucinations the the biggest threat to the success of
0:56
an AI right now and trying to put guard reels around it so
0:59
that you're not giving incorrect information?
1:01
Or is there something else that keeps you up at night?
1:08
This is identity at the center if it has anything to do with
1:13
IAM. This is the go to podcast now
1:18
your hosts Jim McDonald and Jeff Stedman.
1:28
Welcome to the Identity of the Center podcast.
1:30
I'm Jeff, and that's Jim. Hey, Jim.
1:32
Hey, Jeff, how are you? Not so bad yourself.
1:35
I just living the allergy life man.
1:37
Even though we've we've gotten to the point of daylight savings
1:39
time is over and temperatures are getting cooler, my allergies
1:43
just keep getting worse. So if it sounds like I only
1:47
breathe through my mouth, that's why.
1:50
Well, yeah, if you could just not breathe and make sure the
1:52
audio quality sounds good, that would be fantastic for me.
1:55
Yeah sure I'll I'll die mid episode.
1:58
Well, today's special episode because we've got a sponsored
2:00
spotlight and we're joined by our friends from Strabacity.
2:05
His name is Steven Cox, he's a Co founder and CTO.
2:07
Well, welcome here in a second, but just to make it crystal
2:10
clear, this entire episode is sponsored thanks to Stromacity.
2:14
We work with folks like Steven all the time to try and come up
2:17
with different entertaining shows around.
2:19
You know, there are specific viewpoints into the different
2:21
things that they approach from identity space.
2:22
So I'm sure people enjoy it. We've had actually Steven on
2:25
before in the show. Steven, you joined us way back
2:28
in episode 112. So welcome back to the Identity
2:31
of the Center podcast. Thanks a lot guys.
2:33
Yeah, it's great. It's great to be on.
2:34
Always enjoy talking with you guys and, and I'm proud to have
2:37
been #112 on your on your podcast.
2:41
An early supporter, we appreciate it.
2:44
You were with Trivacity then, you're still with Trivacity now,
2:46
and that website is Trivacity dot AI, so make sure you get it
2:49
out there. STRIVACITY dot AI definitely
2:53
encourage to be able to go check it out.
2:55
But why don't we start sort of at the very beginning, you know,
2:58
tell us about Trivacity and give us a refresher on how you got
3:02
into the digital identity space. Yeah.
3:05
So I've been in the digital identity space now for, gosh,
3:08
about 10 years. I came from a sort of network
3:12
and endpoint threat detection background.
3:15
I worked in that space for a number of years.
3:17
I've been in cybersecurity for probably about 20 years now.
3:22
I joined about 10 years ago, I joined a company that your your
3:28
listeners are probably familiar with secure auth with who is now
3:32
my Co founder of Trivacity, Keith Graham.
3:35
And so I focused on, you know, workforce identity management
3:38
and customer identity management at at secure auth, which, you
3:41
know, informed a lot of my, you know, my current beliefs around
3:45
customer identity platforms. So, yeah, that's kind of how I
3:49
got my start on, you know, in, in, in this space.
3:54
And you know, Stervacity is a, is a customer IAM solution.
3:59
We got started in about 2019. So gosh, we've been going about
4:03
5-5 years, 5 1/2 years now. So we're focused on the, the
4:08
customer identity problem end to end, right.
4:10
You hear lots of, you know, lots of terms thrown around this
4:13
space, customer identity, consumer identity, but
4:15
fundamentally like we're ACCIM solution and that's the problem
4:18
we're we're interested in solving for our customers.
4:22
So what is it about your guys's approach to the consumer or
4:26
customer because we're talking Siam, right?
4:28
CIA am here. What makes you guys unique
4:31
because I think that's the question that I always have for
4:33
for folks in the space. OK, well, you know, there's lots
4:35
of solution in the space. How are you guys different and
4:37
how do your customers measure success with the, you know with
4:41
your products? Yeah.
4:43
So there's a couple things I focus on when we talk about
4:46
differentiation, You know what, what sort of, you know, brings
4:49
us out from the pack and and the and the three things I like to
4:52
really talk about are are architecture, our capabilities
4:56
and our focus on on customer experience.
4:58
So in terms of architecture, you know, we spent a lot of our
5:02
early days focused on building a platform that could scale with
5:05
the unique needs of customer identity.
5:08
You know, as we are building, you know, we kind of set some,
5:12
some design principles out to, to help guide us, you know, and
5:15
those principles of have evolved over the years, of course.
5:17
But you know, I can give you a high level kind of what we were
5:19
thinking. You know, one of the, one of the
5:22
big ones was like going after the cloud hesitancy problem,
5:26
right? They're, they're, you know, this
5:28
is less of a problem than it was now than it, than it was in
5:30
2019, but it is still an issue. And it centres around that
5:34
hesitancy of organizations to move certain types of data or
5:38
systems to the cloud for a variety of reasons, right.
5:41
We learned a lot about this in previous experiences before
5:44
servacity. So we wanted to really attack
5:46
that head on. Another thing we sort of
5:50
observed previous experiences was that CIM projects were sort
5:53
of excruciatingly long and laborious, you know, kind of
5:57
often for the wrong reasons. And we wanted to sort of go
6:00
after that problem. You know, customer identity has
6:04
some unique requirements. It requires you to to move
6:07
quickly, to iterate quickly as sort of customer patterns change
6:11
and you discover new customer patterns.
6:14
We wanted to bring the whole customer identity journey under
6:18
the umbrella of a single platform, right?
6:19
Something that had not yet been achieved in 2019 and arguably
6:23
still has not been achieved or only been achieved by us, by us
6:27
today. And you know, sort of sort of
6:30
goes along with that are are things like brand management,
6:34
which was sort of horribly broken until we came along.
6:36
So you have these, you know, you have these orgs that are sort of
6:39
brand of brand organizations, you know, have a sort of a
6:41
parent company and multiple sub brands.
6:44
That's a challenge we run into. So yeah, so we developed an
6:48
architecture that takes these challenge into into account,
6:50
right? We're built on Kubernetes.
6:52
Our platform scales with demand. It can be upgraded using rolling
6:56
upgrades, which means no downtime.
6:59
You know, each of our customers, and this is a big point, gets an
7:02
isolated instance of our product, right?
7:04
We call this isolation by design and that's really what we're
7:08
what we're doing to sort of address that cloud hesitancy
7:12
issue, right. It also helps with sort of data
7:14
sovereignty requirements. But we believe this is sort of
7:16
superior to multi tenant architectures.
7:20
And then orchestration, we have an intrinsic orchestration
7:23
capability that's tailored to the customer journey runs in a
7:26
serverless environment and then everything is protected by a
7:28
sort of role based access control, right?
7:30
So, so, so that's sort of the architecture differentiation in
7:34
terms of capabilities. You know, as I mentioned, we're
7:36
sitting here today, we're kind of the only solution that covers
7:38
the full spectrum of customer identity needs right from a
7:41
single product, single admin console.
7:44
So if I put on my marketing hat, you know, I would like I say
7:46
things like, hey, we can handle customer identity from sort of
7:49
registration to the right to be forgotten, right?
7:51
So the sort of beginning to end. And so you know, our, our
7:55
capabilities cover things like registration, self-service,
7:59
adaptive auth, identity verification, consent
8:02
management, fraud detection, right.
8:05
And then in terms of customer experience, you have to think
8:08
about CIM systems as fundamentally workflow or
8:11
journey systems, right? You, you have to build a whole
8:14
layer on top of that IM tech around generating flexible flow
8:19
based UIS for customers to interact with.
8:21
So we think we've done a pretty good sort of differentiate job
8:24
and sort of making it really easy to do this.
8:27
So how do our customers measure success with the solution?
8:30
It's a great question. I love, I love that question.
8:32
There's, there's kind of two areas that I would focus on
8:36
here. 1 is time to value, right? We, we talked about sort of
8:40
deployments being long. This, you know, this is time to
8:45
value is largely focused on things like deployment, sort of
8:47
set up of the solution, things like customer journeys and sort
8:51
of the migrations of of users, right.
8:54
And the other is sort of ongoing KPIs around things like adoption
8:57
metrics and customer engagement. And this is something where we
9:00
really differentiate with our, with our dashboard, right?
9:05
So, you know, you've got to make that really easy.
9:09
You know, we, we can report on a cup on a few dozen different
9:12
metrics, you know, across things like conversion rate, you know,
9:15
how many visitors are actually converting into registered users
9:18
on your platform, engagement rate, like how are your users?
9:22
Are your user base sort of coming back every month, right?
9:25
How long are they taking to register from the moment they
9:27
click sign up to the moment they're redirected back to the
9:30
customer portal? So these are things that you
9:32
have to report on, right? So that so that they can your
9:35
customers to sort of take those KP is back to their their
9:37
management team and sort of show that they're, they're doing a
9:40
good job, right. So let me ask you a real basic
9:42
question because I feel like sometimes there's just double in
9:46
the details. Is there a difference between
9:49
consumer IAM and customer IAM or they really the the same thing?
9:54
Yeah, there is a difference. I would say it's like a Venn
9:57
diagram, right, where there's where there's overlap and
10:00
there's and there's separate, there's separate capabilities,
10:03
right. So a lot of the sort of
10:04
fundamental IAM technology underneath, you know, things
10:08
like federation, you know, open ID connect SAML, those sort of
10:12
basic, you know, text are, are certain certainly within the
10:17
middle of the Venn diagram between customer identity and
10:20
and, and and workforce identity. But when you get into things
10:23
like sort of customer metrics, you know, like how long is a
10:27
user taking to register? Are you getting people that are
10:30
sort of registering and walking away those there, there's not a
10:35
whole lot of overlap there, right?
10:36
Those are very customer specific metrics and those are things
10:39
that, you know, you're sort of marketing team wants to know
10:43
that your users are converting, right?
10:45
That your users are making it in, in a, in a, in a workforce
10:47
environment, you, you kind of have a, a captive audience,
10:50
right? Like they're, you know, you're
10:52
not going to have somebody that sort of refuses to register with
10:55
the workforce, you know, access management system.
10:58
And so there's a, you know, that there's that sort of where the,
11:01
the sort of the, you know, on the outside of the, of the Venn
11:03
diagram, you know what I mean? But yeah, it's a good question.
11:07
Yeah, Steven. So I thought it was real
11:09
interesting that our, our conversation started going down
11:12
this kind of marketing track because, you know, big thing I
11:15
wanted to talk about today was AI, of course, but even before
11:19
that, it's just the topic of identity security.
11:23
And you know, we had this episode scheduled for several
11:27
months in advance. I've been in communication with
11:31
Keith. And then just recently Gartner
11:35
kind of identified you guys in the, in the paper, it's called
11:40
emerging tech, tech landscape for startups and security
11:43
software. And there is a digital identity
11:48
column, if you will. And I think it's interesting
11:52
because I'm I'm still not sure what is the difference between
11:56
digital identity and identity security.
11:58
I think that's maybe an episode for the future.
12:02
But what do you think? It was about Trivacity that made
12:07
Gardner sit up and take notice. Yeah.
12:10
I mean, so Gardner's been talking about us, you know, for,
12:15
for a while across a couple of different areas, which is kind
12:19
of interesting. Journey time orchestration is 1
12:22
where we've appeared in their research because that's a
12:25
capability of our product, you know, sort of fraud and risk is
12:31
one because we, you know, we have some fraud controls.
12:33
We have things like identity verification and user
12:36
experience, right, which is going to something that's going
12:38
to make us stand out because it's, it hasn't been a focus of
12:41
a lot of our, a lot of our competition.
12:44
And also, you know, we're, we're, we're one of, you know, a
12:47
small number of entrants into this space in the last five
12:50
years that's taken significant investment, right?
12:53
And, and especially 1 that's focused on the, on the customer
12:57
identity problem and has made sort of, you know, customer
13:01
traction, made some significant customer traction into, into
13:04
this space. So I think it's great that
13:07
they're taking notice. I mean, it's helping us with
13:09
brand awareness. It's very crucial for our
13:11
company, our stage to have that. And it's also kind of a sign
13:15
that we're, we're probably getting things right, you know,
13:18
from, from from the analyst point of view, you know, and in
13:22
terms of sort of other, you know, other analysts activity.
13:26
You know, we actually debuted as a leader in the Forster wave for
13:31
CIM in 2022, which is just absolutely unheard of.
13:35
Like for an, A, a start up to sort of debut as a, as a leader,
13:40
you know, we were one of the only three leaders #2 after
13:44
Fordrock at the time. There's another way of doing
13:47
soon, I think. I believe.
13:48
We believe it's going to be sometime in December.
13:51
So you know, Gardner doesn't have a magic quadrant
13:55
specifically for CIAM. But you know, like I mentioned,
13:58
they're talking about us, you know, across other areas of
14:00
research, which is great. And if I'm going to be, you
14:03
know, if I'm going to be self-serving, you know, I'll say
14:06
that, you know, such a young vendor debuting as a leader, you
14:09
know, again on Forrester is just, is really a testament to
14:11
the, to the strategy and just how good our product is.
14:15
You know, I think you know our architecture, our breadth of
14:18
capabilities, our focus on customers experience is really
14:21
what's behind the interest from from the analysts.
14:25
Yeah. So you talked about the
14:26
isolation by design. I think that's just really
14:30
interesting, but talk to us about why that's important from
14:34
a identity security standpoint. Yeah, so I, I came from an
14:40
incident response background. So I, I worked in network and
14:43
endpoint security for a company called Mandiant, probably heard
14:47
of, you know, later acquired by Google and you know, I sort of
14:52
saw the the, the worst of the worst of, of breaches, right.
14:57
So when you know you're asking a, a large org, you know a large
15:02
financial institution, a large healthcare business to sort of
15:06
move their their data to the cloud, you better be protecting
15:10
it, right? So one of our sort of
15:13
fundamental design principles was we're going to, we're going
15:15
to give an isolated instance of our product to each of our
15:18
customers, right? That means that means that if
15:20
one of our customers is breached, the blast radius
15:23
between one customer to the next is, is null, right?
15:27
Like there's no, there's no shared infrastructure, no shared
15:30
keys, databases, services between any two of our
15:34
customers, right? And and that's, that's
15:38
absolutely critical in customer data, customer identity, because
15:41
you're storing, you know, sensitive customer data, you're
15:43
storing passwords, you're storing MFA, you know, MFA
15:47
credentials like this is really, really sensitive data that
15:50
you're storing. So you have to sort of take a
15:52
really strong security, you know, posture with how you
15:57
design these systems. There was a LinkedIn post that
16:00
Jim actually made about why I am tech companies aren't really
16:04
getting into the generative AI space of their products yet.
16:06
I think it's going to be a little bit dated, Jim, no
16:08
offense, because I think we're starting to see it more out
16:10
there. But what?
16:12
And you responded back to that, like what prompted that response
16:16
back? Was it just familiar with Jim or
16:18
you're like, oh, hold on a second buddy, you're way off
16:19
base here. Well, I mean one, I, I do like
16:23
you guys. So I, you know, I, I do.
16:25
We'll get you everywhere on the show.
16:27
Yeah. So, so that was certainly one,
16:30
but I don't know it just it, it was interesting.
16:33
It was, it was during a time when we were sort of deeply
16:35
pondering that very question, right.
16:38
So I sort of struck me as a very relevant question from from Jim,
16:42
kind of considering where we are in the space, where we are in
16:46
sort of the development of AI, you know, where we are in the
16:49
identity space. You know, every sort of product
16:52
leader right now is having to make a decision as to whether
16:56
integrating, you know, generative AI and into their
16:59
platform makes sense, right? And, and, and IAM products are
17:03
no, are no different, you know, than that.
17:05
So that was kind of the reason. Jim, Jim, you wrote this article
17:10
and I don't want to leave people behind.
17:11
So you kind of give like a gist of it so you can catch people
17:14
who are listening or watching. Yeah, it wasn't really so much
17:17
an article as much as it was, you know, a a true question.
17:22
Like it seems like every product is now baking in AI.
17:27
But in the IM space, at least, what I'm seeing is, you know,
17:32
versions of AI are more big data analytics, machine learning.
17:38
They're not generative AI like what you get when you go out to
17:42
open AI or, you know, Gemini or something like that.
17:47
And, you know, I throw this one back to Steven because, you
17:51
know, is that what you see as well?
17:52
Like there's a difference between machine learning and big
17:56
data analytics and AI, right? Yeah, I think, yeah, definitely
18:00
you need to stick, take a step back and and and sort of think
18:03
about like what is the intersection between AI and ML,
18:06
right. Like the media often gets this
18:08
wrong, as you, as you've likely noticed, you know, and I think I
18:12
think the way to position is that, you know, AI is just sort
18:15
of overarching term of which machine learning is a subset,
18:20
right? It's, it's correct to say that
18:22
ML is a form of AI, but ML has been sort of historically
18:26
focused on finding patterns and data and getting smarter over
18:29
time so that you can make decisions on what it finds and
18:32
protects. So if you think about ML, we've
18:36
actually like, we specifically Stervastity has had it in our
18:38
product for a number of years, right?
18:40
Specifically around the behavior analytics problem.
18:44
So let's sort of watching behavior, user behavior
18:46
detecting when there's an anomaly, you know, sort of
18:48
making access decisions based on that.
18:51
And to be honest, like this ML based behavior analytics is, is
18:55
by no means a new technology. The first, the first UEUEBA
18:58
vendors were got their start in what 2014?
19:02
And I mean, that's an eon in, in technology terms.
19:07
So it's, but it's also means it's very tried and true
19:09
technology for, for, for what it's designed for right now.
19:12
If you're talking specifically about generative AI or Gen.
19:15
AI, that's what we sort of most recently added to our platform.
19:19
And what I think a lot of you know, product practitioners are
19:23
sort of pondering right now as to what makes sense for them.
19:28
You know, if you talk about Gen. AI as soaring sort of also being
19:31
a subset of the overall AI space, it also makes use of
19:36
machine learning. Gen.
19:37
Gen. AI models are designed to emit
19:39
or produce data versus simply analyze and, and classify data.
19:45
But they also have other, other elements that sort of fall into
19:48
that bucket. You know, a good example would
19:50
be like natural language processing, right?
19:52
The chatbot aspect of of Gen. AI systems is a is a somewhat
19:57
separate space and it's, you know, it's all about determining
19:59
user intent. Or what they call grounding in
20:04
the in the Gen. AI space.
20:06
So you know, when you sort of couple a prompt that you're
20:10
sending to an LLM with live data, right?
20:13
You, you, you might have heard this referred to as retrieval
20:18
augmented generation or RAG, which is the way that a lot of
20:21
products right now are implementing AI into their
20:25
system. So, you know, we circle back,
20:26
like I say, we've had ML in our product for some time.
20:29
Gen. AI was a was a recent addition.
20:31
I hope that was a pretty clear description of of how I see see
20:34
them differently. No, it's great.
20:36
I mean, I feel like we can, we can really dive into this
20:38
because we kind of joke all the time that this we should turn
20:40
this into like AI at the center of the way the this is moving.
20:44
So my definition of AI changed when I first saw the open AILLM
20:50
sort of released and I was like, oh wow, OK, That AI before that
20:55
really was like machine learning and sort of like behind the
20:58
curtains. And now all of a sudden we've
20:59
got a text based interface at the time and now it's even voice
21:03
that you can, you know, buddy up with and ask questions.
21:06
Now, do you trust it or not? It's different.
21:08
It's a different, you know, answer or a question.
21:10
But did your definition or your thought process around AI change
21:14
when you saw that or you'd already seen that coming?
21:16
I mean. Oh, I mean 100%.
21:20
I mean, the, I was, you know, I was as I was sort of, you know,
21:24
thinking about, you know, putting some notes, some
21:26
thoughts and notes together today for, for the discussion
21:28
today. I, I realized that chat TBT was
21:31
only real was only released like in 2022.
21:34
It was like 2 years ago. I mean, that's crazy that it
21:37
seems like, it seems like it's been like 20 years, you know
21:40
what I mean? And so, yeah, I mean, I think I,
21:43
I think I was just as shocked as as a lot of the folks in, in, in
21:48
the technology space where the sort of power of these of these
21:51
systems of 100%, it completely shook my, my, my understanding
21:57
of the space 100%. So let's get more into that AI
22:01
because you know, I mentioned the website Stravacity dot AI.
22:04
It's literally in the name. What are what are some of the
22:08
things that we can expect if I if I pull up trivacity?
22:11
Give me some examples of like use cases.
22:14
Yeah. So you know, on the benefits,
22:17
you know sort of like the benefits of pulling this into
22:20
our product, there's a couple things that really just line up
22:24
with how we have built the product from the beginning,
22:27
right. What one of one of those is sort
22:30
of lowering the bar, you know, of entry to, to IAM systems,
22:34
right? We, we always say that you kind
22:37
of, you don't need to, you shouldn't have to have a PhD in
22:40
IAM to set up a customer journey, right?
22:43
And so, you know, Gen. AI makes a really good companion
22:48
to work with as you're sort of configuring and maintaining
22:51
your, your IAM systems. You know, things like context
22:55
sensitive help where you're asking questions of an assistant
22:59
on how to, you know, that AI assist on how to do certain
23:02
tasks within the platform. Like, you know, we can tell
23:06
where you are in the admin console.
23:08
We can send along pertinent contacts and our docs to the LLM
23:12
to craft a very specific answer. Like, so hey, how do I set up
23:16
bot detection? Or how do I change my password
23:20
policy to be more restrictive? You know, like we can really
23:23
help with context sensitive help, right?
23:25
That's a, that's a huge, there's one thing we wanted to build for
23:28
a while is constant context sensitive help.
23:31
And Eugenia is a great, a great way, you know, to to sort of
23:36
facilitate that interpretation of data is another one that we
23:42
thought was really important. You know, like I am isn't just
23:45
about like MFA and federation, right?
23:48
We can also generate a lot of data.
23:51
It's often hard for less technical people to crock,
23:53
right? And you know, specifically in in
23:57
customer identity, you know, where some of your users of the
24:01
platform may actually be in fact customer service
24:04
representatives, CSR's who are deeply technical in IAM areas,
24:09
but they might actually be actively on the phone with a
24:11
customer and they're trying to cut, you know, troubleshoot a
24:13
customer issue. So roll ups of these types of
24:17
these log data of this customer journey data can actually be
24:20
really helpful in getting to an answer quickly when a customer's
24:22
unhappy, right? We can say like, oh, you're this
24:25
user is struggling to get through the identity
24:27
verification process or they're not able to finish their
24:30
password enrollment, you know, those kind of things, right?
24:33
So that's really helping there. And then also like one of the
24:36
things we wanted to really focus on and I think is really big
24:38
benefit is, is sort of the guidance towards best practice
24:42
or optimization aspects. So, you know, there's a lot of
24:46
best practice out there, password guidelines, adaptive
24:49
access policies, you know, how you know, what if you could
24:52
guide A-Team or a user towards improving their posture based on
24:56
what they're actually seeing in the product, Specifically on the
25:00
CIAM side, What if you could got a, you know, got a team towards
25:03
optimizing customer experience or improving, you know,
25:06
conversion rates, spice, what you're seeing in the logs, what
25:09
you're seeing in the data, what you're seeing in the, in the
25:11
dashboard. So those are, you know, those
25:14
are kind of those are kind of the focuses and kind of the
25:16
things that you can kind of do with with our, our, our current
25:20
AI assistant in the product. Yeah, I'm kind of I'm interested
25:26
in in, you know, visualizing in my head.
25:29
How does that take place? So that last part that you're
25:32
talking about where it's like, all right, I want the system can
25:37
make suggestions to your client in terms of access policies or
25:44
things. How does that manifest?
25:47
So when the user comes along and they're looking at a particular
25:53
portion of the admin console and they ask a question, we do a
25:59
form of retrieval augmented generation, which is RAG, which
26:02
is just a, a, a really complex way to say that we take the
26:06
prompt, we take the context of what they're looking at in the,
26:10
in the admin console and we take a relevant section of our docs.
26:15
And we send all of that to the LLM at the same time.
26:18
And we say, give me an answer built from this prompt, this
26:22
context, and this documentation. Now this is obviously very, very
26:28
simplified version of a complex thing, but that's what happens.
26:33
And then the LLM has what it's what it needs to answer the
26:37
question in a way that's very specific to our product and
26:41
relevant to the context of what the user is, is, is staring at
26:45
at that at that point in time, right.
26:47
And this is a very, this is a pattern that you will see in the
26:51
Gen. AI space right now of people
26:52
doing these, these rag based systems that are marrying a
26:56
prompt with context, with documentation, with maybe stuff
27:00
on the web, right? Open AI can do that when you
27:03
when you ask a question about current events, it's going to go
27:06
out and search the web, search news articles for relevant stuff
27:10
and then give you a context relevant answer.
27:13
That is another form of rag, right, that that you're seeing
27:16
Open AI implement in their ChatGPT tool.
27:20
So it's a common pattern. That seems like a a really
27:24
clever way to help with the security of the model itself and
27:29
reduce the chance of a hallucination, right?
27:32
I think everyone's kind of concerned about will the AI give
27:34
me a correct answer and can I trust it?
27:36
And what if it comes up with something to put you off the
27:38
wall? But it sounds like if you're
27:39
able to limit the input into that model and have it just be
27:43
based on your documentation, you know the chances of it going off
27:46
the rails are probably a little bit, you know better for it not
27:49
happening. 100% and like, and that, and that goes back to the,
27:54
the, the comment I made about grounding, right, grounding is,
27:58
is, is a, is a necessary technology and sort of
28:01
implementing Gen. AI systems, Gen.
28:03
AI products, because basically what you're doing is you're,
28:07
you're, you're closing those guard rails in a little bit,
28:09
right And you're giving, you're giving an LLM more information
28:12
to give you an accurate answer, right.
28:16
And if you think about like, I love the hallucination topic.
28:19
I, I think it's AI think it's, I think it's great that you
28:21
brought that up. I mean there's, there's been
28:23
some really fun stories about that in the, in the media.
28:27
And I think, I think it's really useful for, for, you know, your,
28:31
your listeners to understand like what a hallucination is
28:35
and, and why you, you need to understand why they happen.
28:39
I, and I think if you, if you look at what an LLM
28:42
fundamentally is, it's, it's, it's a probability machine,
28:45
right? I mean, I'll probably get
28:47
blowback from your users for that simplification, but maybe
28:51
it's better to say that they're sort of a, they're sort of
28:53
fundamentally probabilistic in nature.
28:56
So what they're trying to do is they're trying to figure out
28:58
what the next word, character phrase is based on the context
29:03
that they know about what you, what you've sent it, right?
29:05
So, you know, a hallucination can kind of be thought of as a,
29:10
as a statement that's statistically probable, but
29:12
factually incorrect. I, I, I like that.
29:15
I read that statement somewhere. I was like, that's perfect way
29:17
to describe it. Statistically probable,
29:18
factually incorrect, right. And again, this is one of the
29:21
reasons that people are building grounding systems around AIS
29:25
like like RAG, because it's a safety mechanism, It's a
29:28
security mechanism, like you said, right?
29:30
To make it more, make the answer more reliable, make it more
29:33
accurate. It's certainly a concern for us
29:36
as we were, as we were building the product, building it into
29:39
our product. And it's also why you see
29:41
disclaimers like on chatbot systems.
29:44
It's like, hey, LLMS can make mistakes.
29:46
Please, please verify your answers because hallucinations
29:49
are and will continue to be a problem even with grounding
29:51
systems around around, you know, the, the implementation of
29:55
these. So yeah, I think it's, I think
29:58
it's a great question. Is are hallucinations the the
30:02
biggest threat to the success of an AI right now and trying to
30:06
put guard rails around it so that you're not giving incorrect
30:09
information? Or is there something else that
30:10
keeps you up at night? I mean, there's, you know,
30:13
there's a hallucinations are definitely a challenge.
30:15
I don't, I don't, I think you can kind of especially when
30:19
you're not, you're not sort of implementing what you know, what
30:21
I would consider a public API, a public AI system like like
30:25
ChatGPT. Ours is very specific to our
30:28
product. So we can really control what
30:31
sort of goes into it. But you have all kinds of other
30:34
challenges like, you know, prompt injection is one, right?
30:38
Because if you know, it's not all that different than like a
30:40
sequel injection where you're, you're, you're taking a prompt,
30:44
you're formulating it into a larger prompt and you're
30:47
injecting that in the LLM. And a smart user might figure
30:49
out what the format of your prompt is and sort of break out
30:53
of it in some way, right? And try to get data out of your
30:55
system. But you can also put, you know,
30:58
a system like ours. You can also put guard rails
31:00
further down the road so that you're not able to return or
31:04
you're only talking to certain APIs that are read only.
31:06
And you're, you're, you know, you're, you're putting some
31:08
security guidelines past the LLM point within the product.
31:13
So those kind of things keep me up at night.
31:16
I honestly, I think the other thing that keeps me up at night
31:18
is sort of the, the, the cost around these systems, right?
31:22
That's, that was a, the first thing, the first one of the sort
31:26
of the first things that I thought about when, when we said
31:28
let's, let's put this into our product was like, Oh my gosh,
31:30
how much is this going to cost? Right?
31:32
Like it's compute and memory. It's you requires you to sort of
31:36
have specialized GPU heavy instances in the in the cloud.
31:41
And so you know that when you're sort of talking about large
31:43
scale systems where a user, you know, where you know, you have
31:47
customers that have, you know, 10s of millions of users, you
31:51
know, a large portion of which login every month.
31:53
Like are you are you being careful about how you're
31:56
leveraging, leveraging the tech right with regards to large
32:00
scale deployments? You know, that's that was that
32:03
was sort of a big worry for me, less of a worry now, but but was
32:07
a big worry in the at the at the, you know, the the start.
32:10
Yeah, you're starting to get into where I think the the big
32:14
value and the big threat are. So I think the big value is
32:18
being able to, you know, enable say marketing you brought up in
32:25
the marketing example for business users to query the data
32:29
so they can start saying, OK, well, how many people logged in
32:33
in the last 24 hours? That's one level.
32:36
Another level might be now let's branch out into information
32:40
that's on the Internet. How many people logged in in the
32:42
last 24 hours from somewhere that is was raining, you know,
32:47
and you know, so the, the concern it brings up is like,
32:52
OK, how far can it extend? Could I, you know, what do I
32:57
have to do to create security boundaries to make sure that I
33:01
don't say, all right, what is my competitor doing on this system?
33:05
How many users do they have? And maybe I do that in some kind
33:09
of like less obvious way, right? Maybe I know that my competitor
33:14
is using this system. So I, I start asking questions
33:18
like more at like an industry level, like what do
33:22
organizations that are in my industry, how many users do they
33:26
normally have? Things like how many, how much,
33:29
how many times do people log in the last 24 hours for other
33:32
organizations that are in my industry?
33:34
So, Mike, I think that's the real power would be to be able
33:39
to ask, you know, reporting style questions that it could
33:44
really think about and maybe you could even pitch a scenarios
33:48
like, all right, well, if a hurricane took out power over
33:54
the, you know, this large portion, Jeff and I certainly
33:58
know that problem. How would that, how would that
34:02
impact my logins? I, I knew this as I am, this is,
34:07
you know, I can't really basically go as far as to say
34:09
how would that impact my sales? But maybe, you know, something
34:13
like that. And then I think the, the big
34:17
threat is the more powerful that becomes, the more security
34:22
becomes an issue, which probably goes to like, how did you guys
34:26
go about deploying this? Is this are you leveraging like
34:29
the APIs of some big tech platform?
34:33
Did you develop your own AI or like how did you build AI into
34:37
your product? Yeah.
34:39
So again, we, you know, we we focused on isolation from the
34:47
get go, right, which actually put us in a pretty good position
34:50
for implementing this technology into the platform.
34:54
So we are, we're leveraging an existing LLM, but it's local to
34:59
our customer environment, our individual customer environment.
35:02
So they get their own instance of our LLM within, within the
35:06
customer within their own, you know, product instance.
35:10
So we're not using a public LLM like open AI, right?
35:13
And as soon as you open that door where, you know, these
35:16
systems are able to sort of go out to the Internet or, you
35:20
know, go out to a third party system, you're, you know, that
35:24
opens a lot more, a lot more risk than having a system that's
35:28
sort of self-contained within, within our admin console within
35:31
our own customer instance. It only has access to that
35:33
customer data within within that instance.
35:36
So from in terms of like a, an isolation or sort of a, you
35:41
know, securing a customer perspective, we, we were kind of
35:45
already in a good position for that right in the way we wanted
35:48
it to implement that. And there's, there's a lot of
35:52
local open source models out there that are extremely
35:57
powerful and well suited to RAG systems.
36:01
Like what we like, what we built with Instrabassity, Everything
36:05
else, you know, around it we built is custom, right?
36:08
The actual RAG implementation, you know, the user interface,
36:12
you know, the sort of, you know, AI assist like interface we have
36:16
in our admin console, things like that.
36:17
That's all, that's all custom. But yeah.
36:22
And then to to kind of like agree with the point that I was
36:26
making where like definitely I think the things that you talked
36:29
about a lot of power in that and a lot of benefit in having
36:34
contextual base help. But I think kind of that next
36:39
frontier is, you know, just leaving it open to non-technical
36:44
people to be able to query the the system, query the data.
36:49
Yeah, yeah. And and that's, that's, that is
36:52
something that we can do as well.
36:53
Like you can, you can pop into our account events interface and
36:57
sort of ask like, hey, who are the, you know, tell me about the
37:01
tell me about the users that have logged in today in terms of
37:03
like where did they come from and, and what were they doing?
37:06
And is there anything anomalous, right?
37:08
Like you can do things like that, but it's, you know, it's
37:11
obviously very secured, it's obviously very controlled to
37:14
that specific customer data within that customer instance.
37:18
And we don't sort of allow, you know, you to sort of go like
37:21
allow our system to sort of go out and query the Internet or,
37:24
or sort of other external systems.
37:28
And the other thing that's interesting too, is that, you
37:30
know, we, we have role based access control in the product
37:35
and we treat the LLM like any other subset of functionality
37:40
within, within the product or any other sort of area that lets
37:43
you get access to sensitive information.
37:45
So if you don't want certain employees or groups to have
37:48
access to the LLM, then you don't give them that right in
37:52
the in the admin console and they can't, right.
37:54
So if you want to control who sort of has access to those,
37:58
that's that's a that's a great way to do it is with RMAC.
38:02
That was the question I was going to get into is like how
38:05
you control access to either the LLM or what the LLM has access
38:12
to. I think that that second
38:14
scenario is a lot more difficult, right?
38:17
It it is yeah. And so if you think about these
38:20
sort of rag based AI systems, they're kind of they're kind of
38:23
chains, right? So you, you might, you might
38:26
have multiple LLM calls, you might have multiple API calls,
38:29
right? You might have a, a, a, a first
38:31
date LLM call that's sort of like, Hey, what is this customer
38:34
asking for, right? Then you might make a call out
38:37
to an API, probably one of my, it's probably going to be one of
38:40
our internal APIs to pull data. You might get some context back.
38:44
Then you might make another LLM call with that context to get
38:48
the final answer, right. So they're, they're implemented
38:50
as chains, but that when you implement them as chains like
38:53
that, you also have an opportunity to it, you know,
38:55
sort of introduce security restrictions at different steps,
38:58
right. So it's like, hey, yes, I am not
39:00
going to allow any API calls that do a write or, you know,
39:06
only can access this subset of APIs within our product.
39:10
That that that's kind of the, that's kind of the idea, yeah.
39:13
So Steven, where do you see, where do you see this being five
39:18
years down the road? Like you said, two years since
39:21
Chap GPT kind of hit the mainstream and look where it is
39:27
already. I mean, five years might be, I
39:29
might as well say, you know, in an eternity.
39:32
But where do you see this five years down the road impacting
39:35
the identity space? Well, I mean, if I could, if I
39:38
could send the, the the shrug emoji to your, to your, your
39:42
users, I would. It's it's really hard to say.
39:45
I mean, I think I thought you positioned it really well
39:47
earlier when you asked like was it was it sort of, you know, was
39:50
it sort of a big change on way, the way you think about things
39:53
when, when Chad JP take him out because it it, it, it really
39:56
was. It's really hard to say because
39:58
we're, we're constantly being surprised on what these LMS can
40:02
do. I mean, generating like lifelike
40:04
video and all kinds of, you know, crazy stuff like this and
40:08
that, you know, the AI firms sort of keep bringing out their
40:11
latest and, and greatest models, right?
40:13
And they just keep out doing themselves.
40:16
But I'll tell you one thing that.
40:18
That really piqued my interest recently.
40:21
It's that the, the large tech firms like, you know, like
40:24
Google, Meta, they're, they're, they're seeking out other
40:28
sources of, of energy beyond the, the energy grid, right?
40:33
So they're, they're talking about like obtaining access to
40:35
nuclear energy or building their own nuclear, nuclear energy
40:38
infrastructure. That's just insane to me.
40:40
But it, it tells me two things, right?
40:43
It tells me one, we're taxing the energy grid right now, it's
40:48
not going to keep up with the demand of this of this
40:50
technology, right? And it's also, it's also a hint
40:55
that the those firms are sort of acknowledging that, right?
41:00
It means that they're acknowledging that, you know,
41:02
any efficiencies that they design into these systems or the
41:06
distribution of the compute of these systems, you know, pushing
41:09
it down to devices or, or things like that is not going to, it's
41:12
not going to keep pace with the demand, right?
41:15
So that points, that points to, for me to the technology
41:21
reaching a plateau shortly. Because once you run into issues
41:27
with resource scarcity, you know, it quickly becomes
41:31
geopolitical, right? And you, you know, we, we all
41:34
know, we all know humanity has done a great job at sort of
41:37
sanely sharing resources and it's in its history, right?
41:41
So but yeah, I mean, I kind of think we're only scratching the
41:46
surface of this tech right now. I mean, like I say, ChatGPT
41:49
came, like you've mentioned, came only out and, you know, two
41:52
years ago. Seems like seems like eons.
41:55
I mean, that's nothing. That's, you know, I I would
41:59
equate unleashing eon the world as, you know, equivalent to the
42:03
the advent of the consumer Internet, the invention of the
42:07
microchip, you know, the, the industrial revolution.
42:10
I mean, I think it has that level of impact on the world.
42:13
It's like I I can't imagine where it's going to be in five,
42:17
1050 years from now. You know, I mean, it's, it's
42:20
insane to think. About mind blowing.
42:23
I mean, do you think that it's going to evolve so quickly that
42:29
whatever governance is needed to, you know, keep it from
42:35
becoming a dangerous weapon, that it's going to outpace that
42:41
ability to govern it or that we will be able to govern it?
42:47
And it's kind of a crazy question, but.
42:48
It's, it's a tough question. I, I, I think it will probably
42:53
outpace our ability to, to, to, to govern it, especially with
42:57
the fact that, you know, you have sort of open source
42:59
capabilities around this. I think, you know, governments
43:05
haven't really do done a good job at sort of keeping up with,
43:08
with technology governments in the governance in the past.
43:10
So there probably will be a crackdown at some point.
43:13
But, you know, it just all speaks to more of like keeping,
43:16
keeping the keeping the inventors of this type of
43:19
technology involved in the, in the governance of this
43:22
technology. But it's, it's, it's, it's going
43:24
to be a very tough problem. I think it's really going to, I
43:26
think we're, it's going to outpace our ability to sort of
43:28
keep a, keep a lid on it. Yeah, to get this through as any
43:31
guide, it's difficult to keep up with tech.
43:33
It just moves too quickly. Yep, Yep, yeah.
43:35
And I mean, and our elected officials aren't really aren't
43:38
really sort of equipped to, to understand this technology at
43:42
any level and, and sort of, you know, the the applications of it
43:45
and the impacts of it. They'll get there, they'll get
43:47
there. But it's, it's moving so fast
43:49
right now, you know? Going back to the power
43:52
requirements, sounds like we need a a Tony Stark type person
43:54
to come up with an arc reactor that we can stick into a data
43:58
center somewhere and have it just power that thing and just
44:01
be good. Yeah, I mean, I just, I to me,
44:03
it was just insane to see the the tech firms sort of going
44:05
after nuclear power to to keep their data centers running.
44:09
You know what I mean? It's just crazy.
44:12
Microsoft's and, you know, Microsoft AI and you know,
44:15
Microsoft nuclear like that is a that is a weird thought to say,
44:19
yeah, all right, why don't we go ahead and wrap it up there?
44:22
This has been, I mean, we could talk AI all day for sure.
44:26
I'm looking at this Trivacity dot AI website and you, you, you
44:29
almost stole all the kind of the Thunder of the website because
44:31
on the screen I'm looking at right now, you know, it's, it
44:33
has like this little picture of access control, you know,
44:38
settings and you got the AI assistance.
44:40
I, you know, basically asking, hey, are these settings secure?
44:42
Which I think is a really cool way to approach securing the
44:45
product, right? Especially you mentioned, you
44:47
know, being able to train it on your documentation.
44:48
It's like, OK, here's exactly what I'm looking for.
44:51
And I love that idea of leveraging those AI assistants
44:55
to help with that because I don't want to pick up the phone
44:58
and talk to somebody. I want to do it myself, right?
45:00
I want to be able to look it up and say, OK, click this, do this
45:03
here. OK, good.
45:04
Now I kind of get it. So it's a really cool example of
45:06
how how you guys are approaching it.
45:07
You've got it. So let's talk a little bit about
45:11
astrophotography because I, I did my first astrophotography
45:17
just a few weeks ago. Oh, yeah.
45:19
So I'm in the Asheville, NC area and we had Northern Lights
45:25
basically come down as far down as we were.
45:27
And that's the first time I've ever experienced anything like
45:29
that. And it was almost invisible to,
45:32
you know, the naked eye. But when I pulled out my phone
45:34
and that's all I had to kind of take pictures.
45:37
All of a sudden I've got this picture and it's like Reds and
45:41
purples are in the sky and it's, you know, 2 in the morning or
45:44
whatever it is. I'm like, wait a second.
45:46
I like, that's not what I saw, but that's what the camera
45:49
picked up. And I'm just fascinated by this
45:51
idea. So I started looking at it like,
45:52
well, can I download an app or something to kind of, you know,
45:55
capture this better than, than just the, you know, the iPhone?
45:58
And so I'm curious because I think you're into the
46:01
astrophotography stuff too. I I want to get into this a
46:03
little bit more with you. Yeah, yeah.
46:06
So, you know, it's, it's actually something that I
46:08
recently got into as well, like as in this year.
46:11
And it's, you know, it's something I've I've I've wanted
46:14
to, I've had a I've had a telescope for a while, like a
46:16
nice one, but I've never gotten into sort of the like post
46:20
processing part of it, you know. And so I think one of the things
46:24
that was sort of holding me up on it was, you know, some of
46:29
these rigs that you build for sort of capturing deep space
46:32
objects can be like incredibly expensive, like, you know, five,
46:35
$10,000 if you really want to build like a really nice rig,
46:38
But there's actually this really cool and something you might
46:41
want to look into if you have an interest.
46:43
There's really there's these really cool set of like smart
46:46
telescopes coming out where you can.
46:49
They're not very big. They, you can sort of put them
46:53
out, you know, in your yard and, and train them on a, on a, on an
46:57
object that you want to record. And then they have an app where
46:59
you sort of control it and, and capture images.
47:02
And they actually can produce some pretty amazing images.
47:06
So they'll take exposures, like they'll take like 10 second
47:08
exposures and you know, you leave them out there and maybe
47:11
over the night you take 510,000, you know, 10 second exposures.
47:16
And then they have stacking tools, right?
47:19
So you take these, all of these images, you stack them together.
47:23
You know, it does some crazy calculations on the pixels to
47:25
decide what is noise and what is, what is image, you know,
47:30
what is, what is signal And, and you get this beautiful image of
47:34
like a deep space object, right? Like, so like of a, you know, of
47:36
a nebula or you know, of a, of a Galaxy, right.
47:41
And I, I just, I was, like I said, I was afraid from the cost
47:45
requirements and I was afraid that I wouldn't like the post
47:48
processing angle of it. Just, you know, because it's,
47:51
it's, it's a whole new sort of technology to learn, but I, I
47:54
love it. It's, it's actually another
47:57
place that the AI has made an appearance because you can, they
48:02
have these sort of AI driven noise algorithms that are that,
48:06
that people have developed where they try to figure out like what
48:09
is signal and what is noise and sort of remove the noise from an
48:12
image. So you can, you know, you can
48:13
produce these visually brilliant images from a little telescope
48:17
sitting out on your backyard because of, because of the way
48:20
you're stacking the images and the way that you're, you're,
48:22
you're denoising them, right. So yeah, it's, I've actually, I,
48:27
you know, I've spent like it's been a really, we've had a
48:29
really run good run of the sort of clear nights here in
48:31
Virginia. And I've been out, I've been out
48:33
in my backyard like almost every night doing, you know, taking
48:36
pictures, so. Well, that's the key, right, is
48:39
light pollution kind of really impacts that.
48:41
So if you're in a in a rather remote area where there's less
48:44
of the light pollution you're, you're more prone to be having
48:47
to have better quality for the images, right?
48:50
Yeah. And they they call that the,
48:52
they call that the Bortle scale. And so it goes from, I think it
48:56
goes from like 1:00 to 9:00, I think.
48:58
But I'm in a Bortle 5 area. So I'm right in the middle
49:03
because I'm right outside of a, of a major airport where I live.
49:06
So that you get a lot of light pollution from, but we actually,
49:09
we go down to the Outer Banks. I'm North Carolina a lot I'm
49:12
sure you're familiar with. And a lot of people don't
49:15
realize that is actually one of especially the Ocracoke areas,
49:18
one of the darkest areas in the United States.
49:20
And it is one of the darkest areas on the, on the East Coast
49:22
as well. So you can actually get Bortle
49:24
2, which is the second darkest on the Bortle scale.
49:27
So I and it having the little smart scope, you can, you can
49:31
kind of take it down there. It's not, it doesn't require a
49:33
lot of space. So like when we go down there,
49:35
I'll, I'll, I'll take it down there and sort of put it out on
49:37
the back deck and, and get some really good pictures because
49:39
it's so dark down there. So from what I understand of
49:42
Ocracoke Island, you take a ferry to get there you have to
49:46
have a 4 by 4. You drive on the beach and pick
49:50
your spot and you can camp or whatever.
49:54
A lot of people go surf fishing. So you go onto this island and
49:59
you set up your telescope and your camera.
50:02
Yeah, Yeah, you can do that, Yeah.
50:03
I mean, Ocracoke, you have to take a ferry to Yeah, but there
50:06
actually is a Route 12 goes. It's a, you know, a major road
50:09
goes all the way down to the city of Ocracoke.
50:11
But yes, if you go out on the beaches of Ocracoke when it's a
50:15
clear night, you can see the, the Milky Way without any
50:20
imaging equipment at all. And it's it's amazing.
50:23
Ocracoke's beautiful. If you haven't been there, I
50:25
highly recommend it. So that's.
50:27
That's what our ancestors did, right?
50:29
They just looked up in the sky. There's no light pollution.
50:32
Yeah. Yeah, We did a episode pretty
50:36
recently with a guy named Nitin. He's the CTO over at Civil
50:43
Security and he rents part. He's in a group of people who
50:49
rent AA telescope and Jeff made the the statement, which I
50:55
thought was spot on. Like there's supposed to be an
50:57
expensive telescope if it takes several people to kind of pull
51:01
their money. To bring a trailer and
51:03
everything. Oh my gosh, Yeah, yeah, yeah.
51:05
Some of some of the, the, the Dobsonian telescopes are really
51:09
big. Yeah.
51:10
So they're, and they're actually less, less expensive.
51:13
I don't know why that is, but but yeah, they can be pretty
51:16
big. But yeah, one of the reasons I
51:18
got into the sort of smart telescope area was because it's
51:21
they're portable. You know what I mean?
51:23
You can, it can, it can pack down into a little box and I can
51:26
just throw it in the back of my truck on my way down, you know,
51:28
to the Outer Banks. So I had.
51:30
To look up the Bortle scale because I hadn't heard of that
51:33
before, because I know I live in an area that's got less Royce
51:35
provision and it looks like the Asheville area is around a three
51:37
or a four. Wow.
51:39
Yeah, pretty. Good.
51:39
And I've got a mountain range between where my house is and
51:42
the rest of Asheville, so it's even darker, so I feel like I'm
51:45
in a good spot. So if I, yeah, if I drive a
51:47
little bit West of where I am, I can get to the Shenandoah
51:50
Mountains and I can get some really good, really good dark,
51:53
dark spots out there. But, but again, like even in a,
51:57
even in a highlight pollution area, you can pull a lot of the
51:59
light pollution out of the images in the post processing
52:01
part of of this. So there's people that take
52:04
great pictures like out of Tokyo where it's portable 9, you know,
52:08
So it's, you know, some of the denoising algorithms are, are
52:12
pretty fascinating. You stack enough images you can
52:14
you can figure out what's what's noise, what's light pollution
52:17
and what's what's image you know.
52:19
Of course you get those cool like motion ones across the sky
52:22
right where it's. Like yeah, with the starting.
52:23
Image and here's the only thing that's moved, right?
52:25
Yeah. And stuff like that.
52:26
Yeah. So if someone wanted to get into
52:28
this, what would be the beginner?
52:30
And let's say it's me on the on my back deck trying to have
52:35
maybe just a slightly better set up to do some astrophotography.
52:39
So, so the, the, there's a company called CWO and they have
52:45
a, they have a couple of scopes called the sea stars.
52:49
And so there's like AC * S 30 and AC star S50.
52:52
And they are the, the, the highest quality, sort of easiest
52:57
beginner scopes that, that you can find.
53:00
They'll actually livestack with, they'll livestack the, they can
53:03
livestack the, the images that you take while you're using the
53:06
app. So you don't even really need to
53:08
do post processing on your own, unless you want to throw
53:11
something into like Photoshop or something.
53:13
They'll do the stacking for you. So you pop the scope out in the,
53:17
in the, you know, wherever you want to, you want an image, you
53:19
tell it, you actually can tell it, like what deep space object
53:22
or what the moon or, or what planet you want to look at.
53:26
It'll point the scope at that, it'll start taking pictures,
53:28
it'll stack it for you and you get a nice little image on your
53:30
phone and your, or your tablet. And so I'm a big fan, big fan of
53:35
that company. What do you do with the pictures
53:37
that you take? I post them on Instagram and and
53:41
and Facebook and, you know, share them with share them with
53:45
my buddies. Wallpapers.
53:46
Print them out. Put them on behind you.
53:48
Yeah, well, maybe I haven't done that yet, but that that would be
53:50
a good idea. You know, my kids also enjoy
53:53
seeing what I'm doing and taking pictures off, so I share it with
53:57
them too. Well, maybe Sea Star wants to
54:00
sponsor an episode in the future.
54:02
There you go. Yeah, stuff like that.
54:03
They'll appreciate the mention. It's a tiny little.
54:06
I'm looking at the one right now and it's definitely a tiny
54:08
little thing. It's much smaller than I
54:09
anticipated it being. Yeah.
54:10
Well, that's super cool. Which I think is probably a good
54:14
spot to leave it with a product mentioned for another company.
54:17
So let's get back to stravacity. Stravacity dot AI Steven, you,
54:22
you've been great again. You know, I hope we can talk
54:24
again in the near future. I really like what you guys have
54:28
done with the the AI assist. And so I'm definitely encourage
54:30
people to go check it out. And I've been a big fan of what
54:33
you guys been working on for the last, you know, couple years
54:35
now. So it's been exciting to see
54:36
that evolution, you know, happen over and I won't say it real
54:40
time because, you know, a couple years is is a blink in the eye
54:42
of a tech spot, but it's been really impressive.
54:46
Thanks, I and I and I always appreciate talking to you guys.
54:48
I hear, I hear that you guys are like celebrities at conferences
54:51
now and people walk up to you and recognize you and stuff.
54:54
So I'm, I'm, I'm honored that you, you would have me on your
54:57
on your show again. So I'm looking forward to the
55:00
next time. Well, we appreciate that.
55:02
Yeah, for sure. And definitely appreciate all
55:05
the wishers to come up and and I try not to be awkward when I can
55:08
do it. All right, let's go and leave it
55:11
there for this week. You can find Strabasi on the web
55:14
again, Strabasi dot AI. We'll have links in our show
55:17
notes and people can check that out.
55:18
And then Steven, I'll put a link to your LinkedIn profile as well
55:22
in case people have questions or maybe just want to talk about
55:24
astrophotography. I'm sure, I'm sure you would
55:26
appreciate either of those things.
55:28
And as far as we go, we're on the web, idacpodcast.com and we
55:34
got the YouTube channel going now, idacpodcast.tv.
55:37
So feel free to like and subscribe.
55:38
Jim's shaking his head. Like make sure I get that plug
55:40
out. But yeah, thanks for everyone
55:42
for listening and are watching and we'll talk with you all in
55:45
the next one. You've been listening to
55:49
Identity at the Center. We hope you've enjoyed the show.
55:53
Make sure to like, rate and review, and we'll be back soon.
55:58
But in the meantime, hit the website at
56:00
identity@thecenter.com. See you next time on Identity at
56:06
the Center.