October 17, 2024 • 17 mins
Join host Jim Jockle in an enlightening episode as he sits down with Kanishk Parashar, co-founder and CEO of Powder, a pioneer in AI-powered solutions for wealth management. Together they delve into how Powder’s technology is streamlining processes like document analysis, cutting task times by as much as 90%, and pushing the boundaries of operational efficiency. Kanishk also shares insights into the safeguards for AI-driven accuracy and what this could mean for the future of financial services. If you're curious about how AI is transforming the industry, this is an episode you won’t want to miss.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:06):
Welcome to Trading Tomorrow navigating trends in
capital markets the podcastwhere we deep dive into
technologies reshaping the worldof capital markets.
I'm your host, jim Jockle, aveteran of the finance industry
with a passion for thecomplexities of financial
technologies and market trends.
In each episode, we'll explorethe cutting-edge trends, tools
and strategies driving today'sfinancial landscapes and paving
(00:29):
the way for the future.
With the finance industry at apivotal point, influenced by
groundbreaking innovations, it'smore crucial than ever to
understand how thesetechnological advancements
interact with market dynamics.
Welcome to Trading Tomorrownavigating trends in capital
(00:54):
markets.
For this episode, we're lookingat generative AI agents and how
they're changing the waycompanies operate.
By automating time-consumingtasks, these advanced
technologies are enablingbusinesses to place their focus
elsewhere.
Genai has seen a significantsurge in adoption and investment
over the past year.
According to a recent McKinseysurvey, 65% of organizations are
(01:17):
now regularly using GenAI,nearly doubling the amount from
a report that they ran 10 monthsearlier.
For this episode, we're excitedto be joined by Kanishk Parashar
, the co-founder and CEO ofPowder, a leading company in the
AI space that is at theforefront of this technological
transformation.
The former co-founder and CEOof RCI Navigator, kanishk
(01:37):
previously worked at CoinInc.
As the founder and CEO Powder'sinnovative solutions use Gen AI
to help wealth managers workfaster and more accurately.
Join us as we discuss theirjourney, the impact of their
technology and how companieslike Powder are leveraging
cutting-edge advancements totransform business operations
and stay ahead in a competitivelandscape.
(02:00):
Kanishk, thanks so much forjoining the pod.
Speaker 2 (02:02):
Thank you for having me here, James.
Speaker 1 (02:03):
And why don't we just start with a brief overview of
Powder and your services?
But let me just say this isprobably our 30th podcast and
Powder is the best name of acompany we've had on yet, so if
you could wrote that into youroverview, that'd be great to
hear.
Speaker 2 (02:18):
Yeah, absolutely yeah .
So I'll give you an idea of howwe came up with Powder.
We were at our last company.
It was called Navigator, we'rethe same founding team and
Navigator was acquired byAdapart in January 2021.
Now what happened was that, aswe were working with tens of
different RIA firms, what werealized was that a lot of
advisors they told us that theirdifferentiator was not their
(02:41):
investing expertise, but it wastheir ability to build trusted
relationship with their clients.
And that realization, you know,also led to the fact that the
information needed to buildthese trusted relationships
they're spread out in notes,conversations, documents you
know, we call it unstructureddata and there's a lot of manual
effort required to gather andleverage that information.
So now, with the launch ofgenerative AI, which is actually
(03:04):
great and making sense of allthis unstructured data, you know
, we felt like we can bring in anew technology that can help
advisors make sense of this datawith a lot more speed and
accuracy, you know.
So, after chatting with a bunchof firms, we narrowed it down
to our focus.
You know, powder, which isdocument analysis for a state,
you know, and the name is justcompletely random.
(03:25):
It's the easiest one to pickand it allows us to move forward
with two syllables.
Speaker 1 (03:31):
Fair enough.
I was wondering if there was adry powder angle in there
somewhere.
I wasn't sure.
Speaker 2 (03:37):
Yeah, actually you're right.
There's actually a dry powderangle in there, but most people
think of it as like skiing, so Idon't mention it too much.
Speaker 1 (03:46):
Fair enough, all right.
Well, I guess I'm the curiousone.
So what are the tasks that yourAI agents are automating, and
is it just a tackle from aproductivity standpoint?
And if yes, you know how muchtime is being saved?
Speaker 2 (04:02):
Absolutely so.
Our first agent is able to reada brokerage document and in
detail, like it breaks out thetax lots.
It fills in missing information, such as if it's missing the
tickers.
It enriches the data.
Like it assigns its assetclasses, and that extract is
ready to use for a firm.
Our customers are telling usthat they roughly save about 90%
(04:23):
of the time they spend on thisactivity and, even more
importantly, they don't everwant to go back to doing it
manually.
Once they use our system,they're happy getting rid of
this mind-numbing task fromtheir lives.
Speaker 1 (04:35):
So I'm an avid AI user, right?
I use ChatGPT, microsoftCopilot to protect work and
sensitive data, you know, buttell me with, with the
availability of all this othergenerative AI, um, you know what
is what is making your companyunique?
Speaker 2 (04:53):
Yeah, you know, um, we're applying generative AI
towards operational efficiencyrather than investing alpha.
So manual work at the firms aremostly done on unstructured
information that's coming inlike notes and so on, right, and
filling out forms.
You know finding things on theinternet, right.
(05:14):
So we believe like being ableto leverage all of your
unstructured data if it's comingfrom a database right, that's
going to be a superpower.
So the operational efficiencyis what our aim is.
Speaker 1 (05:26):
And hallucinations.
Everybody talks about them.
I think the best comment I'veheard on hallucinations is it's
really the AI learning orexploring its own thoughts.
But how would you ensure thereliability and accuracy of the
AI's outputs, especially withcritical business processes?
Speaker 2 (05:49):
Yeah, since we have started to build this technology
, we have noticed the same thing.
There's hallucinations, youknow.
So we have built manyhallucination safeguards we call
it, you know such as we havethe AI read the statements twice
and then we compare the outputs, For example.
We also do simple checks on top, such as, for example, for a
(06:12):
tax lot, what is cost basis plusgain?
Does that add up to the assetvalue?
And so, just like this, we dotens of more checks to make sure
the output is accurate andtrustable.
Speaker 1 (06:23):
So how much time are you spent doing training at this
point?
Speaker 2 (06:27):
Believe it or not, that we're using, you know, a
base level LLM.
You know everything we're doingis built on top, you know.
Now we have explored fine tuneLLMs and we have explored uh rag
implementations as well, youknow, but we haven't put that
into like general production yet.
Speaker 1 (06:43):
Got it, got it All right, right?
Well, it sounds like excitingthings are to come, but we'll
dive into that.
So you know where you are today.
What kind of feedback have youreceived from your clients that
have uh integrate powder intotheir operations?
Speaker 2 (06:54):
so the firm's employees.
You know the firms that arecustomers.
Their employees?
You know they're.
They're high-end trained staff.
You know they get paid sixdigits salaries and they say
they don't ever want to go backto doing the eye-watering manual
labor of extracting informationfrom PDFs.
And the deal is now they canfocus on higher value tasks,
(07:16):
such as for pinpointing.
Their value add to a prospect,which is when you get to receive
documents and they're movingfaster with less effort.
Speaker 1 (07:24):
And so you know I'd be remiss if I didn't ask.
You know, obviously there's alot of emerging financial
regulation around the use of AI.
Obviously, data privacy hasbeen at the forefront of
regulatory authorities and withbroad, sweeping regulations put
(07:46):
in place, like GDPR, et cetera.
So how are you addressing dataprivacy and security concerns
when deploying AI solutions?
Speaker 2 (07:54):
Yeah.
So my viewpoint on this is thatgenerative AI is no different
than using cloud solutions,which are already widely
deployed at most firms, you know, in the US and worldwide.
So, to that end, we're close toacquiring a SOC 2 compliance,
which is the industry standardthat everyone trusts and relies
on.
Speaker 1 (08:14):
Oh, that's excellent.
That's excellent.
And so you know, what would yousay?
You know, in terms of the mainbenefits for generative AI for
financial institutions, is itproductivity at this point or,
you know, is it going to movebeyond that, or what do you see
the benefits?
Speaker 2 (08:34):
Yeah, I think you hit it on the head.
It's all about operationalefficiency, right?
Instead of going to search forinformation or figuring
something out, get an answer Wayfaster, way more convenient, as
long as it's accurate, so youknow.
So today you know, for example,it's saving a bit of time and
effort.
It's pretty good.
Tomorrow it's going to be.
(08:55):
It's going to finish morecomplex tasks, tasks that
require multiple steps, you know, and so an AI that enables your
employees to be more efficientis a game changer.
Speaker 1 (09:06):
Obviously, you know, there's the fear of AI taking
over the world and everybody'sgoing to lose their job.
What are your thoughts andreactions to not necessarily
naysayers, but those who aresharing these kinds of concerns?
Speaker 2 (09:22):
A new technology has downsides these kinds of
concerns and new technology hasdownsides, you know and like.
So, for example, one of thedownsides of generative AI is
that it's very, very good atbeing wrong.
It'll confidently tell you towalk off a cliff.
And so I think financialinstitutions, you know, need to
build and partner with companiesthat have built a series of
(09:45):
safeguards which make the AItransparent, you know, and
that's the way to getcomfortable with it.
Speaker 1 (09:50):
Okay, and so you know , as companies are making these
investments, you know howimportant is it for these
institutions to invest ineducation and training for these
workforces, or is it?
Or you know, or I've seenstudies where people are just
turning on new functionalitywith an AI without any really
(10:12):
guideposts or education.
I mean, where do you think thatimportance level should be?
Speaker 2 (10:19):
So there's a quote from the NVIDIA CEO, jensen,
that rings true, which is AI isnot going to replace you.
Someone using AI is going toreplace you.
You know, and the employees youknow of major institutions need
to demand that they geteducated on this new technology
that's coming at them.
You know, and it's urgent thatthe culture is built around
(10:40):
using AI, like starting today.
You know, and this way, they'reready.
When there's a massive leap ofintelligence, they're ready to
move along with it and ifunprepared, they might be left
with horses when there's a roadfull of cars.
Speaker 1 (10:54):
I like that analogy.
I like that analogy.
You know, one of thefascinating I was speaking at a
panel about a month ago and oneof the fascinating quotes.
We were talking marketing andAI and changes in evolution, and
there was this study that wasspeaking that the best users of
(11:15):
generative AI are women over 40as compared to men, because they
, you know, ask detailedquestions and, from a
communications standpoint, youknow, are very exploratory in
the way they speak as comparedto men, who are very curt.
But the converse wasinteresting in terms of it was
(11:36):
women over 40 who were not usingand getting education on AI
tools, where it was more youngermen and younger women who were
more excited.
So I just think it was justinteresting how the best cohort
is the ones who probably needthe most upskilling at this
point or, you know, willingnessto change their ways of working,
(11:56):
which is crazy.
But let me come back to adifferent question.
Right?
So you know, in terms ofchallenges for financial
institutions, right so, in termsof challenges, for financial
(12:16):
institutions.
Speaker 2 (12:17):
what challenges are they facing when integrating Gen
AI, either on their own orthrough other third-party type
solutions?
Yeah, I think the challenge isthat there's a lot of stuff
going on and it's hard to figureout what's really AI, what's
really important and things likethat.
So my inkling here is that tostart with a very specific use
case, that the institution canbuild its comfort experience and
(12:37):
a level of culture, and thenunderstand the positive effects
and also understand thedownsides, and then leverage
that to jump into more use cases.
Speaker 1 (12:46):
As many companies, especially financial
institutions.
A lot of them are behind theGartner hype cycle, if you will,
in terms of adoption ofdifferent technologies.
A lot of them are behind theGartner hype cycle, if you will,
in terms of adoption ofdifferent technologies.
What advice would you give tothese types of financial
institutions as they'rebeginning their AI journey?
Speaker 2 (13:05):
It's simply around the fact that AI seems mystical
if you're not a technologist andyou really have to hone down
and find one specific thingwhere you want to test it out
and understand how it makes adifference.
And building that experienceand understanding the positive
and the negative effects of itbecause there are some issues
(13:26):
you have to address along theway will help the institution
and its employees understand howto leverage it.
What are the real solutions outthere?
What solutions are nice to haveversus, you know, important to
have today and then leveragethat to jump into more broader
use cases?
Speaker 1 (13:43):
so, you know, let's come back to powder for a minute
.
Um, you, you teased a couple.
Uh, you know, maybe, perhapsfuture developments.
But you know, what futuredevelopments or advancements can
we expect from powder in thegen ai space, going forward as
you take over the fintech space?
Speaker 2 (14:02):
Yeah, you know our customers are guiding us towards
our next big agent that we'remaking.
We'll actually launch a meetingnote taker that has vertical AI
specifically built for wealthadvisors.
So what I'll do is it'llautomatically create a rich
client profile that's based onconversations that the note
taker is being a part of, youknow, and the idea is that it'll
(14:25):
capture specific informationthat a human note taker can miss
easily, you know, and they'rekind of sprinkled into
conversations such as, like youknow, clients like interest,
their relationships, theirvalues, their goals and more
things that the advisor wants tocapture, and what this will do
is that later on, an advisor canthen recall any of this
(14:47):
information simply by asking aquestion in the chatbot
assistant, and what we hope isthat the advisors can create a
very complete and thoroughprofile and get to really know
their clients so they can do abetter job of servicing them.
Speaker 1 (15:01):
Well, so you dropped the term and I just want to dig
into that a little bit.
You said the term vertical AI.
Can you explain that to ourlisteners?
Speaker 2 (15:11):
Absolutely so.
A vertical AI goes in depthinto one domain and provides a
very specific use case that istailored to that domain, which
you won't find in a generic AIcapability.
So a generic note taker willnot help you understand a client
the way an advisor wants tounderstand a client.
Speaker 1 (15:32):
And so I just want to stay here for a second, because
this is very different.
And you know, I've had oneconversation where an individual
was talking about assigningpredetermined personas, you know
, to a particular AI.
You know, are you doing any ofthat to make the note-taking
(15:52):
even more specific to the enduser, which is that financial
advisor?
Speaker 2 (15:57):
We're not doing that yet, but that's a great idea.
Oh, okay, that's a wonderfulidea.
Yeah, and I'll send you a bill.
Speaker 1 (16:05):
I hate to say it, but we've reached the final
question of the podcast.
We call it the trend drop.
It's like a desert islandquestion.
So if you could only watch ortrack one trend in Gen AI, what
would it be?
Speaker 2 (16:19):
Yeah, you know it's like the whole industry in
general, like open source.
Llms are becoming better andbetter.
You know they're becomingcheaper and cheaper.
So today, you know we seehundreds of companies launching
super amazing use cases, butthose companies still have
highly skilled engineers astheir employees.
It's expensive.
In the coming years, thelowered cost and technology
(16:41):
leaps, I think we'll be able toturn everyone on the planet into
an engineer.
So if you can think of it, thenthe LLMs can make it for you.
So I can't imagine what happenswhen the entire human race can
be a skilled engineer.
There's no limit.
Speaker 1 (16:59):
Oh, I'm going to develop the next Candy Crush.
That's all I know.
So I want to thank you so muchfor your time and you know,
congratulations on Powder, andyou know wishing you the most
continued success and we'lldefinitely be keeping you on the
radar.
Speaker 2 (17:14):
Thank you, James, Really really happy to be here
and thank you very much.
Speaker 1 (17:25):
Thanks so much for listening to today's episode and
if you're enjoying TradingTomorrow, navigating trends and
capital markets, be sure to like, subscribe and share, and we'll
see you next time.
Welcome to Trading Tomorrow navigating trends in
capital markets the podcastwhere we deep dive into
technologies reshaping the worldof capital markets.
I'm your host, jim Jockle, aveteran of the finance industry
with a passion for thecomplexities of financial
technologies and market trends.
In each episode, we'll explorethe cutting-edge trends, tools
and strategies driving today'sfinancial landscapes and paving
(00:29):
the way for the future.
With the finance industry at apivotal point, influenced by
groundbreaking innovations, it'smore crucial than ever to
understand how thesetechnological advancements
interact with market dynamics.
Welcome to Trading Tomorrownavigating trends in capital
(00:54):
markets.
For this episode, we're lookingat generative AI agents and how
they're changing the waycompanies operate.
By automating time-consumingtasks, these advanced
technologies are enablingbusinesses to place their focus
elsewhere.
Genai has seen a significantsurge in adoption and investment
over the past year.
According to a recent McKinseysurvey, 65% of organizations are
(01:17):
now regularly using GenAI,nearly doubling the amount from
a report that they ran 10 monthsearlier.
For this episode, we're excitedto be joined by Kanishk Parashar
, the co-founder and CEO ofPowder, a leading company in the
AI space that is at theforefront of this technological
transformation.
The former co-founder and CEOof RCI Navigator, kanishk
(01:37):
previously worked at CoinInc.
As the founder and CEO Powder'sinnovative solutions use Gen AI
to help wealth managers workfaster and more accurately.
Join us as we discuss theirjourney, the impact of their
technology and how companieslike Powder are leveraging
cutting-edge advancements totransform business operations
and stay ahead in a competitivelandscape.
(02:00):
Kanishk, thanks so much forjoining the pod.
Speaker 2 (02:02):
Thank you for having me here, James.
Speaker 1 (02:03):
And why don't we just start with a brief overview of
Powder and your services?
But let me just say this isprobably our 30th podcast and
Powder is the best name of acompany we've had on yet, so if
you could wrote that into youroverview, that'd be great to
hear.
Speaker 2 (02:18):
Yeah, absolutely yeah .
So I'll give you an idea of howwe came up with Powder.
We were at our last company.
It was called Navigator, we'rethe same founding team and
Navigator was acquired byAdapart in January 2021.
Now what happened was that, aswe were working with tens of
different RIA firms, what werealized was that a lot of
advisors they told us that theirdifferentiator was not their
(02:41):
investing expertise, but it wastheir ability to build trusted
relationship with their clients.
And that realization, you know,also led to the fact that the
information needed to buildthese trusted relationships
they're spread out in notes,conversations, documents you
know, we call it unstructureddata and there's a lot of manual
effort required to gather andleverage that information.
So now, with the launch ofgenerative AI, which is actually
(03:04):
great and making sense of allthis unstructured data, you know
, we felt like we can bring in anew technology that can help
advisors make sense of this datawith a lot more speed and
accuracy, you know.
So, after chatting with a bunchof firms, we narrowed it down
to our focus.
You know, powder, which isdocument analysis for a state,
you know, and the name is justcompletely random.
(03:25):
It's the easiest one to pickand it allows us to move forward
with two syllables.
Speaker 1 (03:31):
Fair enough.
I was wondering if there was adry powder angle in there
somewhere.
I wasn't sure.
Speaker 2 (03:37):
Yeah, actually you're right.
There's actually a dry powderangle in there, but most people
think of it as like skiing, so Idon't mention it too much.
Speaker 1 (03:46):
Fair enough, all right.
Well, I guess I'm the curiousone.
So what are the tasks that yourAI agents are automating, and
is it just a tackle from aproductivity standpoint?
And if yes, you know how muchtime is being saved?
Speaker 2 (04:02):
Absolutely so.
Our first agent is able to reada brokerage document and in
detail, like it breaks out thetax lots.
It fills in missing information, such as if it's missing the
tickers.
It enriches the data.
Like it assigns its assetclasses, and that extract is
ready to use for a firm.
Our customers are telling usthat they roughly save about 90%
(04:23):
of the time they spend on thisactivity and, even more
importantly, they don't everwant to go back to doing it
manually.
Once they use our system,they're happy getting rid of
this mind-numbing task fromtheir lives.
Speaker 1 (04:35):
So I'm an avid AI user, right?
I use ChatGPT, microsoftCopilot to protect work and
sensitive data, you know, buttell me with, with the
availability of all this othergenerative AI, um, you know what
is what is making your companyunique?
Speaker 2 (04:53):
Yeah, you know, um, we're applying generative AI
towards operational efficiencyrather than investing alpha.
So manual work at the firms aremostly done on unstructured
information that's coming inlike notes and so on, right, and
filling out forms.
You know finding things on theinternet, right.
(05:14):
So we believe like being ableto leverage all of your
unstructured data if it's comingfrom a database right, that's
going to be a superpower.
So the operational efficiencyis what our aim is.
Speaker 1 (05:26):
And hallucinations.
Everybody talks about them.
I think the best comment I'veheard on hallucinations is it's
really the AI learning orexploring its own thoughts.
But how would you ensure thereliability and accuracy of the
AI's outputs, especially withcritical business processes?
Speaker 2 (05:49):
Yeah, since we have started to build this technology
, we have noticed the same thing.
There's hallucinations, youknow.
So we have built manyhallucination safeguards we call
it, you know such as we havethe AI read the statements twice
and then we compare the outputs, For example.
We also do simple checks on top, such as, for example, for a
(06:12):
tax lot, what is cost basis plusgain?
Does that add up to the assetvalue?
And so, just like this, we dotens of more checks to make sure
the output is accurate andtrustable.
Speaker 1 (06:23):
So how much time are you spent doing training at this
point?
Speaker 2 (06:27):
Believe it or not, that we're using, you know, a
base level LLM.
You know everything we're doingis built on top, you know.
Now we have explored fine tuneLLMs and we have explored uh rag
implementations as well, youknow, but we haven't put that
into like general production yet.
Speaker 1 (06:43):
Got it, got it All right, right?
Well, it sounds like excitingthings are to come, but we'll
dive into that.
So you know where you are today.
What kind of feedback have youreceived from your clients that
have uh integrate powder intotheir operations?
Speaker 2 (06:54):
so the firm's employees.
You know the firms that arecustomers.
Their employees?
You know they're.
They're high-end trained staff.
You know they get paid sixdigits salaries and they say
they don't ever want to go backto doing the eye-watering manual
labor of extracting informationfrom PDFs.
And the deal is now they canfocus on higher value tasks,
(07:16):
such as for pinpointing.
Their value add to a prospect,which is when you get to receive
documents and they're movingfaster with less effort.
Speaker 1 (07:24):
And so you know I'd be remiss if I didn't ask.
You know, obviously there's alot of emerging financial
regulation around the use of AI.
Obviously, data privacy hasbeen at the forefront of
regulatory authorities and withbroad, sweeping regulations put
(07:46):
in place, like GDPR, et cetera.
So how are you addressing dataprivacy and security concerns
when deploying AI solutions?
Speaker 2 (07:54):
Yeah.
So my viewpoint on this is thatgenerative AI is no different
than using cloud solutions,which are already widely
deployed at most firms, you know, in the US and worldwide.
So, to that end, we're close toacquiring a SOC 2 compliance,
which is the industry standardthat everyone trusts and relies
on.
Speaker 1 (08:14):
Oh, that's excellent.
That's excellent.
And so you know, what would yousay?
You know, in terms of the mainbenefits for generative AI for
financial institutions, is itproductivity at this point or,
you know, is it going to movebeyond that, or what do you see
the benefits?
Speaker 2 (08:34):
Yeah, I think you hit it on the head.
It's all about operationalefficiency, right?
Instead of going to search forinformation or figuring
something out, get an answer Wayfaster, way more convenient, as
long as it's accurate, so youknow.
So today you know, for example,it's saving a bit of time and
effort.
It's pretty good.
Tomorrow it's going to be.
(08:55):
It's going to finish morecomplex tasks, tasks that
require multiple steps, you know, and so an AI that enables your
employees to be more efficientis a game changer.
Speaker 1 (09:06):
Obviously, you know, there's the fear of AI taking
over the world and everybody'sgoing to lose their job.
What are your thoughts andreactions to not necessarily
naysayers, but those who aresharing these kinds of concerns?
Speaker 2 (09:22):
A new technology has downsides these kinds of
concerns and new technology hasdownsides, you know and like.
So, for example, one of thedownsides of generative AI is
that it's very, very good atbeing wrong.
It'll confidently tell you towalk off a cliff.
And so I think financialinstitutions, you know, need to
build and partner with companiesthat have built a series of
(09:45):
safeguards which make the AItransparent, you know, and
that's the way to getcomfortable with it.
Speaker 1 (09:50):
Okay, and so you know , as companies are making these
investments, you know howimportant is it for these
institutions to invest ineducation and training for these
workforces, or is it?
Or you know, or I've seenstudies where people are just
turning on new functionalitywith an AI without any really
(10:12):
guideposts or education.
I mean, where do you think thatimportance level should be?
Speaker 2 (10:19):
So there's a quote from the NVIDIA CEO, jensen,
that rings true, which is AI isnot going to replace you.
Someone using AI is going toreplace you.
You know, and the employees youknow of major institutions need
to demand that they geteducated on this new technology
that's coming at them.
You know, and it's urgent thatthe culture is built around
(10:40):
using AI, like starting today.
You know, and this way, they'reready.
When there's a massive leap ofintelligence, they're ready to
move along with it and ifunprepared, they might be left
with horses when there's a roadfull of cars.
Speaker 1 (10:54):
I like that analogy.
I like that analogy.
You know, one of thefascinating I was speaking at a
panel about a month ago and oneof the fascinating quotes.
We were talking marketing andAI and changes in evolution, and
there was this study that wasspeaking that the best users of
(11:15):
generative AI are women over 40as compared to men, because they
, you know, ask detailedquestions and, from a
communications standpoint, youknow, are very exploratory in
the way they speak as comparedto men, who are very curt.
But the converse wasinteresting in terms of it was
(11:36):
women over 40 who were not usingand getting education on AI
tools, where it was more youngermen and younger women who were
more excited.
So I just think it was justinteresting how the best cohort
is the ones who probably needthe most upskilling at this
point or, you know, willingnessto change their ways of working,
(11:56):
which is crazy.
But let me come back to adifferent question.
Right?
So you know, in terms ofchallenges for financial
institutions, right so, in termsof challenges, for financial
(12:16):
institutions.
Speaker 2 (12:17):
what challenges are they facing when integrating Gen
AI, either on their own orthrough other third-party type
solutions?
Yeah, I think the challenge isthat there's a lot of stuff
going on and it's hard to figureout what's really AI, what's
really important and things likethat.
So my inkling here is that tostart with a very specific use
case, that the institution canbuild its comfort experience and
(12:37):
a level of culture, and thenunderstand the positive effects
and also understand thedownsides, and then leverage
that to jump into more use cases.
Speaker 1 (12:46):
As many companies, especially financial
institutions.
A lot of them are behind theGartner hype cycle, if you will,
in terms of adoption ofdifferent technologies.
A lot of them are behind theGartner hype cycle, if you will,
in terms of adoption ofdifferent technologies.
What advice would you give tothese types of financial
institutions as they'rebeginning their AI journey?
Speaker 2 (13:05):
It's simply around the fact that AI seems mystical
if you're not a technologist andyou really have to hone down
and find one specific thingwhere you want to test it out
and understand how it makes adifference.
And building that experienceand understanding the positive
and the negative effects of itbecause there are some issues
(13:26):
you have to address along theway will help the institution
and its employees understand howto leverage it.
What are the real solutions outthere?
What solutions are nice to haveversus, you know, important to
have today and then leveragethat to jump into more broader
use cases?
Speaker 1 (13:43):
so, you know, let's come back to powder for a minute
.
Um, you, you teased a couple.
Uh, you know, maybe, perhapsfuture developments.
But you know, what futuredevelopments or advancements can
we expect from powder in thegen ai space, going forward as
you take over the fintech space?
Speaker 2 (14:02):
Yeah, you know our customers are guiding us towards
our next big agent that we'remaking.
We'll actually launch a meetingnote taker that has vertical AI
specifically built for wealthadvisors.
So what I'll do is it'llautomatically create a rich
client profile that's based onconversations that the note
taker is being a part of, youknow, and the idea is that it'll
(14:25):
capture specific informationthat a human note taker can miss
easily, you know, and they'rekind of sprinkled into
conversations such as, like youknow, clients like interest,
their relationships, theirvalues, their goals and more
things that the advisor wants tocapture, and what this will do
is that later on, an advisor canthen recall any of this
(14:47):
information simply by asking aquestion in the chatbot
assistant, and what we hope isthat the advisors can create a
very complete and thoroughprofile and get to really know
their clients so they can do abetter job of servicing them.
Speaker 1 (15:01):
Well, so you dropped the term and I just want to dig
into that a little bit.
You said the term vertical AI.
Can you explain that to ourlisteners?
Speaker 2 (15:11):
Absolutely so.
A vertical AI goes in depthinto one domain and provides a
very specific use case that istailored to that domain, which
you won't find in a generic AIcapability.
So a generic note taker willnot help you understand a client
the way an advisor wants tounderstand a client.
Speaker 1 (15:32):
And so I just want to stay here for a second, because
this is very different.
And you know, I've had oneconversation where an individual
was talking about assigningpredetermined personas, you know
, to a particular AI.
You know, are you doing any ofthat to make the note-taking
(15:52):
even more specific to the enduser, which is that financial
advisor?
Speaker 2 (15:57):
We're not doing that yet, but that's a great idea.
Oh, okay, that's a wonderfulidea.
Yeah, and I'll send you a bill.
Speaker 1 (16:05):
I hate to say it, but we've reached the final
question of the podcast.
We call it the trend drop.
It's like a desert islandquestion.
So if you could only watch ortrack one trend in Gen AI, what
would it be?
Speaker 2 (16:19):
Yeah, you know it's like the whole industry in
general, like open source.
Llms are becoming better andbetter.
You know they're becomingcheaper and cheaper.
So today, you know we seehundreds of companies launching
super amazing use cases, butthose companies still have
highly skilled engineers astheir employees.
It's expensive.
In the coming years, thelowered cost and technology
(16:41):
leaps, I think we'll be able toturn everyone on the planet into
an engineer.
So if you can think of it, thenthe LLMs can make it for you.
So I can't imagine what happenswhen the entire human race can
be a skilled engineer.
There's no limit.
Speaker 1 (16:59):
Oh, I'm going to develop the next Candy Crush.
That's all I know.
So I want to thank you so muchfor your time and you know,
congratulations on Powder, andyou know wishing you the most
continued success and we'lldefinitely be keeping you on the
radar.
Speaker 2 (17:14):
Thank you, James, Really really happy to be here
and thank you very much.
Speaker 1 (17:25):
Thanks so much for listening to today's episode and
if you're enjoying TradingTomorrow, navigating trends and
capital markets, be sure to like, subscribe and share, and we'll
see you next time.
Popular Podcasts
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com