April 13, 2024 • 49 mins
Welcome to today's episode where we dive deep into the transformative impact of artificial intelligence on the senior housing investment landscape. Joining us is Kyle Gardner, the COO of NIC MAP Vision, who brings a wealth of expertise to our discussion. Today, we'll uncover how AI is revolutionizing the field, turning intricate market analysis into a streamlined and efficient process essential for investors, operators, and developers across the spectrum—from quaint small towns to bustling metropolitan areas.
With the help of NIC MAP Vision's cutting-edge technology, we'll explore how what once took days of financial analysis can now be accomplished in mere minutes, enhancing the precision and efficiency of investment teams. We’ll also delve into some fascinating niche applications of AI, such as fall prevention in senior housing, illustrating how these technologies are not only improving operational efficiencies but also boosting customer engagement.
Amid a backdrop of rising occupancies and lagging development, we'll discuss how the baby boomer generation is fueling a surge in demand, creating a market teeming with opportunities. Kyle will help us navigate through the latest data trends that suggest we are on the cusp of a golden age for investment in this sector.
It's an exhilarating time to be involved in senior housing investments, and with Kyle's insights, you'll be equipped to lead the way in this dynamic field. So, tune in as we explore the critical role AI is playing in shaping the future of senior housing investments. Join us for a compelling conversation on the cutting edge of senior housing real estate technology.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:07):
so when you go and ask the model to do something,
say, hey, chat, gbt, give me arecipe for turkey dinner.
It goes into its memory to saygreat, I've seen thousands or
hundreds of thousands of recipesand let me find ones that I
understand are connected tocooking turkey or something like
(00:28):
that, and I'll give you anoutput.
Well, for senior housing itmight be.
Hey, nai, go and read theregulations for assisted living
in the state of North Carolinaand then tell me what are the
licensing requirements or thestaffing requirements for
medical professionals.
(00:49):
And so it'll go into its memorybanks or go into a reference
document, consume thatinformation and then produce you
an answer.
Speaker 2 (01:04):
Welcome to the Senior Housing Investors Podcast.
If you are an owner operator,investor, developer or buyer of
senior housing, you've come tothe right place.
The best way to stay connectedwith us is to sign up for our
weekly newsletter athavenseniorinvestmentscom.
This podcast doesn't existwithout you, our community.
(01:26):
Thank you for listening andreach out to us anytime.
Speaker 1 (01:36):
Welcome back everyone .
Today, john Haber is speakingwith Kyle Gardner, the Chief
Operating Officer at NICMAPVision.
They have an interestingconversation about data and
decision-making and how AI isplaying a role in that.
Speaker 2 (01:50):
John.
Thanks, Kelsey.
Today, I have the pleasure ofsitting down with Kyle Gardner,
the Chief Operating Officer ofNICMAP Vision.
He was also Chief Operatingofficer of Vision LTC, which was
acquired by Nick in 2021.
His current responsibility isproviding market analysis tools
(02:12):
and comprehensive data toclients in the senior housing
sector.
Welcome to the show, Kyle.
Speaker 1 (02:18):
Yeah, thanks for having me on John Nice to be
with you.
Speaker 2 (02:21):
Yeah, absolutely, man .
You and I go back four yearswhen you were with Vision LTC
and I called out of the blue andsaid my understanding is you
have an amazing tool, and sotell us a little bit about the
origins of both Vision LTC andthen, as you got acquired by
Nick, tell us that story.
Speaker 1 (02:40):
Yeah, absolutely, nick tell us that story?
Yeah, absolutely.
So I joined Vision LTC back in2017, when it was a small but
promising startup and our goalwas to provide kind of market
analytics tools for investorsand operators and developers
focused on the senior housingspace.
(03:01):
On the senior housing space,there was another product in the
market that was much bigger andmore well-known than us, known
as NICMAP, which provided marketfundamentals, inventory rate,
occupancy absorption metrics and, on the Vision LTC side of the
space, what we found is a lot ofour customers were also NICMAP
(03:21):
customers, so it was a reallynice compliment to each other.
You know the ketchup and hotdogs kind of combination, and so
we were growing the Vision LTCbusiness and then had an
opportunity to join forces withNick, and so we created a deal
where Nick purchased Vision LTCand spun us out into a separate
(03:46):
entity that's now known asNickMap Vision, where the
product suite of what wasformerly Vision LTC and formerly
NickMap is all under one roof,and we've spent hundreds or
(04:15):
maybe even thousands of manhours and a lot of money
combining the products into oneplatform psychographics,
inventory rate, occupancymigration patterns of seniors,
healthcare utilization withinsenior housing communities, the
relationship betweenconstruction activity and
(04:35):
changes in market demand and alot more position ourselves to
serve any stakeholder who has aninterest in the real estate of
senior housing, whether you'reinvesting in it, building it or
operating it, or even selling toit In some cases.
We have data to help you andrecently we've started expanding
(05:01):
into AI as a data provider.
It makes sense and we'll talkabout that, I think, here in a
few minutes.
But we've also got some toolsaround what I call a value add
for our customers.
We have a listings platformwhere customers can buy, sell
and trade senior housing assets.
That's no cost to them, no costto the broker, no cost to the
(05:23):
seller or the buyer.
You just have to be a client ofthe platform to get access.
We've got some healthcare datato help operators become aware
of who the healthcare providersare in the space to artificial
intelligence, specificallythrough OpenAI's GPT models, to
(05:47):
bring AI-powered automations tosenior housing investors and
operators.
So we have a very wide productfootprint, but we focus solely
on the senior housing space andour customers have proven with
their wallets and their feedbackthat we're on the right track.
Speaker 2 (06:05):
So Well, I can say that when we started with Vision
LTC as a company at HavenSenior Investments, we were just
blown away.
And then doing the combinationof both Nick and Vision's data
sets combined has given us theability to really advise our
(06:27):
clients wisely with data andalso on the brokerage side of
our business, being able tounderstand the markets all the
way down to the smallest marketsin the United States, even the
small towns.
It is a beautiful package thatyou all have, and I'm very
excited to learn more about theAI side of the business.
(06:49):
And so you know how does AIplay a role in allowing
operators and investors toaccess data for decision-making,
and I'd love to understand whatthe challenges are currently
and how this AI addresses thosechallenges.
Speaker 1 (07:07):
Yeah, yeah, absolutely.
So to answer that question, letme start first with kind of the
challenges at large that ourdata and our tools solve,
because AI is really anextension of that and a
continuation of that Mission'sthe wrong word, but kind of the
service that we play.
(07:27):
So if you're a real estateinvestor whether it be a
first-time fund entering seniorhousing or you're a REIT
purpose-built for this industryyou need information in order to
make investment decisions right.
So 20 years ago the way to getthat information was to hire a
(07:49):
consultant or a broker to go andput boots on the ground and
maybe live in the market for afew days and do some recon.
You'd probably spend some timethere as well, look at Census
Bureau data and so on, and thenthey'd bring you a large report
and you'd kind of debrief withthem over time and kind of
fast-forwarding to now.
The capability of analyticstools and demography and other
(08:14):
analytics sets like KnickknackVision and like GIS technologies
have enabled kind of desktopstudies where you don't have to
leave the office to get at leastlike a fingertip feel of the
population size and what's inthe market, and Google Maps
gives you some pile of visuals.
You're still putting boots onthe ground before you write the
(08:36):
check.
But the time it takes to accessthat information has gone from
maybe weeks and months to days,and now with AI and the amount
of data that we have in our tool, it can go down to minutes.
And so the challenges thatwe're helping solve is really
helping investors make informeddecisions with more confidence
(08:58):
and in less time, like that'sthe motto that we're kind of
telling ourselves here atNickmap Visions.
That's the motto that we'rekind of telling ourselves here
at Nickmap Visions how do wehelp people execute faster to
either get to an informed no andwalk away from a bad deal, get
to an informed yes and takeadvantage of an opportunity for
their competitors and maybe evenhelp them define what a good
(09:19):
deal or a bad deal is for them.
That's going to changedepending on the shop and risk
preferences and things of thatnature.
But information and intelligenceis kind of at the core of all
we do.
We've been servicing customersfor 20 years on that theory,
going back to the early days ofNICMAP, and now that AI is here,
it just allows us to acceleratethat process even faster, but
(09:44):
it also lets us go considerablydeeper into detail, right?
So I think everyone's familiarwith Census Bureau data or
Bureau of Labor Statistics data.
You know tables on tables ofnumbers and you know CSV
spreadsheets nothing really sexy, but it's the source of truth
that we have.
People are probably familiarwith like Claritas and Esri
(10:07):
those are two data partners thatwe use in our systems and that
data is available.
It's available very quickly,especially in systems like
NICMAP Vision, but it requiresyou as the analyst to make an
opinion.
You know, read the data andthen figure out is this good,
bad, is this normal, is it anoutlier?
(10:30):
Right?
With AI combined with kind ofindustry experience, we can help
the customer go frominformation to opinion or go
from information torecommendation very quickly, and
so AI is kind of an extensionof the you know data as a
(10:52):
service.
Our CEO likes to use the phraseyou know, ai is the arms and
legs and the data is the body.
You know, so it helps make itmakes it more actionable Awesome
makes it more actionableAwesome.
Speaker 2 (11:11):
And to tell our audience a little bit about just
overall artificial intelligenceI'm sure many know what it
means or what it is or just givea high level overview of
artificial intelligence usingOpenAI, chatgpt 4, and here
shortly ChatGPT5 is coming out.
So pretty cool stuff.
So if you could describe kindof in a layman's term what you
(11:35):
mean by AI, yeah, absolutely so.
Speaker 1 (11:39):
Well, sam Altman, ceo of OpenAI, if you're listening,
please release GPT-5 soonerthan later.
But what is AI?
What is a large language model?
I think the average person'sheard about ChatGPT, at least in
passing, since it was releaseda few years ago.
At its core, ai, or largelanguage AI, is a type of
(12:04):
technology.
Under that umbrella or underthat segment of technology,
there's a couple of differentapplications.
There's large language models,machine learning, natural
language processing and a fewother technology applications,
and I am not on the technicalside of the business so I won't
(12:26):
try to explain the nuances ofall of those.
But the way that I understandit, the way that I bring to my
clients and working with our AIengineering team they're the
true geniuses here is largelanguage models is basically a
statistical model that you'vetrained an application to read
(12:48):
and consume information.
So whether that's a pressrelease, a blog article, the
content on a webpage, if youwere to take a picture of a menu
you know at a restaurant andyou could convert that picture
to text, you know the languagemodel could read that text and
consume it.
And then there's kind of amethod where the technology is
(13:20):
reading the text that's on thescreen or on the page and
putting looking at, you know,open AIs GPT-4 is currently, you
know, in April of 24, whenwe're recording this is
currently considered kind of themarket leader in model
performance.
There's other models fromAnthropic and Meta and Google,
(13:41):
but what they're all doing isthey're being trained on a
certain type of data set andthen those models take what
they've learned and, much like ahuman, try to apply that to
complete certain practices orcomplete certain outcomes or
create certain outcomes.
And I know this is probably alittle confusing here, so let me
(14:03):
tie this back together tosomething that real-world
example that makes sense.
So OpenAI goes and trains thesemodels on billions of data
points webpages, pictures,videos.
Just think about the content onthe internet, and OpenAI has
probably seen it or touched itin some capacity newspapers,
(14:24):
blogs, so on and so forth.
As it's reading all thiscontent, it's gathering
information in terms ofknowledge or facts, but it's
also gathering information interms of how the human language
is transcribed and written.
So when you go and ask themodel to do something, say
(14:47):
written.
So when you go and ask themodel to do something, say, hey,
chatgpt, give me a recipe for aturkey dinner.
It goes into its memory to saygreat, I've seen thousands or
hundreds of thousands of recipes, and let me find ones that I
understand are connected tocooking turkey or something like
that, and I'll give you anoutput.
Well, for senior housing andapproach you know where it
(15:08):
becomes applicable for investorsand operators, it might be.
Hey, nai, go and read theregulations for assisted living
in the state of North Carolinaand then tell me what are the
licensing requirements or thestaffing requirements for
medical professionals in thestate, what's the ratio I need
(15:31):
of a med tech or an RN or an LPNto certain residents or any of
those positions even require, soon and so forth?
And so it'll go into its memorybanks or go into a reference
document, consume thatinformation and then produce you
an answer.
It's incredibly powerful, it'sincredibly complex, but at the
(15:51):
end of the day, it's a type oftechnology, albeit it's a
cutting edge one.
So it feels new, it feels, youknow, in some cases kind of
scary, but it's code at the endof the day.
Speaker 2 (16:06):
Well, it's interesting that you spoke about
.
You know, tell ChatGPT for aquestion and it'll go into its
memory and bring out output.
Well, I did that before theshow.
I wanted to quote on whatChatGPT for would come up with
in regards to AI itself, andthey came up with an optimistic
(16:27):
vision, ethical consideration,transformational change but I
like this one the best, and it'sa quote from Stephen Hawkins,
and AI will be the best or worstthing ever for humanity, so
let's make it the best.
And it's so true.
It's really the individualsthat are training these AI
models.
Are you going to train it to bethe best it can be and be the
(16:50):
best for humanity, or are yougoing to create it in a fashion
that is detrimental to us?
I believe in many other beliefsthat it's one of the most
transformational changes thatwe're going to encounter in the
history of the world, and solet's get back to where AI can
(17:11):
add value to the senior housingecosystem today.
Can you riff on that a littlebit?
Speaker 1 (17:18):
Yeah, absolutely.
If you look at the jobs and Idon't mean like job titles or
the positions that people have,I literally mean like the work
to be done by people who workfor a private equity company or
an investment manager or anoperators management company
(17:42):
Look at the job content thatthey're doing every day.
Yeah, if you had to boil itdown to really simple concepts,
are they reading, writing, doingmath?
A lot of work that is done inthe investment workflow and the
asset management workflow comesdown to reading something,
(18:07):
reading it once and committingit to memory, or coming back to
it multiple times, applyingcritical thinking and then
creating an action plan andgoing from there.
Or it's read something, go andbuild a model in Excel to
(18:27):
represent the world and do somescenario analysis of what could
or might happen, and then createan action plan from there, and
that's a very oversimplified wayof thinking about the world.
But when you look at the jobsto be done or the work to be
done by professionals in thisindustry and you realize we
spend a lot of time readingmanagement contracts,
(18:48):
regulations, laws.
We spend a lot of time lookingat offering them randoms, broker
opinions of value, pitch booksand pitch decks and things of
that nature, and then multipleparties are asked to make in
most cases million dollars,sometimes tens of million
dollars decisions on thatinformation.
(19:12):
But what is AI really reallygood at?
It is exceptional at reading,consuming information and
applying a kind of a criticalthinking to that information,
with some caveats.
There has been research overthe last year that, like Chat,
gbt or gbt4 those are the samemodels, by the way that gbt4 can
(19:37):
pass the bar exam thatpracticing attorneys have to
take in the us at a level higherthan the average lawyer.
Yeah right, because it's your,it's reading and critical
thinking and and analysis, butit's really bad at math.
Llms are terrible at math.
So it's your, it's reading andcritical thinking and and
analysis, but it's really bad atmath.
(19:58):
Llms are terrible at math, soit's kind of funny that it can't
do third grade math properly,but it can pass the bar exam,
which is, in certain circles,right Considered kind of one of
the highest white collar jobs ormost prestigiouscollar jobs
that we have in the US.
And so if we play to AIstrengths, it can help us
accelerate our decision-makingtime in certain areas.
(20:22):
It can help us dig deeper intosubject knowledge.
It can help us do scenarioanalysis in many different ways,
and if I was going to breakthose out into a couple of
different examples, let's lookat an investment analyst or a VP
of investment at a firm.
They're reading OMs, they'rereading contracts, they're
(20:46):
drafting OMs, they're reviewingBOVs all to make a decision.
Well, with NICMAT Vision, we'vetaken AI and our deep industry
expertise and createdautomations that use a mixture
of LLMs, like an open AI, not achat GBT or, sorry, a GBT4.
(21:07):
Combine that with our kind ofown proprietary technology is
where, like our engineering team, our developers have created
code-based workflows that arewhat I'll call rigid, meaning
that they're not llm based, thatone plus one is always two, uh.
(21:28):
So we've combined kind of thepower of that custom technology
with the creativity of the LLMsto focus on specific automations
.
And so one of the automationsthat our investors have been
using and love is take aoffering memorandum, just upload
(21:48):
it in any readable format sothe PDF has to be readable or it
can be a Word document.
We'll use the LLM to read theoffering memorandum, pull out
the information that'sinteresting and relevant, based
on some customizable promptsthat our team's created, and
(22:09):
then draft kind of a two page,three page executive summary and
as part of that summary thatthe automation's creating, we're
injecting NICMAP vision datainto it.
So your broker sends you an OM,you load it into the tool.
You answer a couple of basicquestions what's the name of the
(22:31):
subject property?
What do you want to ask theoffering memorandum?
The same way, you kind of entera prompt in the chat GPT, and
then a couple of other basicparameters.
I think there's maybe fivequestions total.
You hit run and in five to 10minutes you know the time it
takes to go refill your coffee.
(22:52):
You come back and you've got athree-page executive summary of
the offering memorandum.
Looking at the subject property, it's going to analyze any of
the financial information that'sin there.
It's going to add market compsfrom our NICBAT Vision market
fundamentals data set.
It's going to pull indemographics specific to your
market area.
It'll even give you questionsthat you should send back to the
(23:14):
broker or to the seller so youcan dig deeper into the product
or, sorry, into the investmentopportunity and kind of go from
there.
Is that automation going toreplace your investment team?
No, absolutely, absolutely not.
It's going to replace yourinvestment team.
(23:36):
No, absolutely, absolutely not.
It's going to make them a wholehell of a lot faster at what
they do, though, because now,instead of waiting, a
well-staffed analyst team who'sputting in long hours maybe
charge that same analysis aroundin two business days.
Maybe say, an average teammaybe takes a whole week.
I'm talking about like a true,like deep dive into the, the OM,
(24:00):
the underlying data, uh, doingtheir own third party analysis.
You know they maybe takes thema week, and you're getting that
same kind of V1 draft in 10minutes.
So imagine what your team could.
You know that same team who ispumping out reports in two to
five days.
You've just given them thatwhole.
(24:21):
You know, let's call it 48hours back minimum, to then go
even deeper or to go look atmore deals, and what could that
do for your, your process?
What could that do for yourprocess?
What could that do for yourcompetitiveness?
I think it can do a lot.
There's other things we're doingas well, just at a high level.
(24:43):
We've created automations tohelp with reviewing income
statements and readingfinancials, providing summary
analyses of those, basicallyhelping you stay ahead of market
trends.
We've got some tools built onthe regulations, obviously
senior housings, regulated atthe state level by a large,
(25:04):
specifically for assisted livingand memory care.
So the regional and nationalproviders have a lot of rule
books that they have to keephandy and, fortunately, ai is
quite good at analyzing those,and we've created some workflows
that, if every building isgetting surveyed and if you're
(25:26):
getting deficiencies, back atsome point, you're going to get
asked to create a plan ofcorrections.
Back at some point, you'regoing to get asked to create a
plan of corrections.
Well, why not have AI do yourfirst draft and then have your
legal experts spend their timevetting the information and
updating the workflow to make itabsolute and applicable to the
(25:46):
reuse case?
But instead of spending theirtime drafting what is basically
a template form, let's put AI towork there.
Or, if you've got questions ofthe regulations, instead of
spending hours and hours goingthrough these terrible websites,
the hundreds of thousands ofclick-through links that the
(26:08):
states love to utilize, why notjust ask a question of a chat
bot and get an answer back inless than a minute?
It's quite powerful.
So everything we're doing withAI really ties back to
accelerating yourdecision-making process and
giving you quick, strategicmarket insights.
Ai is not going to replace yourjob today, and if someone's
(26:33):
telling you that they have notspent enough time using AI, but
if your team is using AI, youwill outperform a team who is
not plain and simple.
Speaker 2 (26:44):
Agreed Bottom line is I always felt that AI was going
to be a companion to us ashuman beings, and that just
makes everything more powerfulour ability to how much they're
walking or how much they aren'twalking.
Maybe that decision makingthrough AI can alert us in the
(27:19):
future to areas that we have noidea we would have uncovered if
we didn't have AI next to us.
So tell us about the future ofthe senior housing ecosystem
when it comes to AI.
What's the future thinkingthat's going on behind the
scenes at NICMAT Vision?
Speaker 1 (27:38):
Yeah, absolutely.
I'll give you a two-part answeron this.
I'll talk a little bit aboutwhat we're working on and how
we're thinking about things.
But I feel like I'd be doingthe pot of disservice, john, if
I didn't share a little bitabout what I'm already seeing in
the market from other providersin the AI space and kind of
(27:58):
where the industry is going atlarge.
I feel kind of unique andprivileged to get to see quite a
lot from the different NICconferences and stuff that I've
been to and it's quiteimpressive and stuff that I've
been to and it's quiteimpressive.
So for NICMEM Vision, we see AIas being a I like your word
(28:19):
like a partner.
We see it as being an enablerfor team members to accelerate
their workflow.
The most value to be had rightnow in 2024 is for research and
customer service and labormanagement, asset management,
investment management workflows.
(28:40):
So whether that might besomeone in the FP&A department,
that might be someone on theacquisitions team at a REIT, it
might be the asset manager of aninvestment fund, it could be
the VP of HR, vp of people at anoperator or something of that
nature.
Those are the areas that we seeit kind of living in right now
(29:02):
and it kind of goes back to whatI was talking about before with
.
It's very good at reading andcritical thinking and doing some
scenario analysis.
And doing some scenarioanalysis when it's going, is it
kind of leading us to think thatthere's potential to help with,
maybe, labor management orlabor assistance at the
community level.
At the management level, thatwe're probably going to see that
(29:26):
in the shape of staffingoptimization.
We'll probably see that interms of embedded features in
current applications, whetheryou're EHR, erp.
I think there's some very clearopportunity there.
The ecosystem at large that'snot something NCHMAP Vision's
(29:47):
likely to want to touch on, butwe see that as kind of an
opportunity for the investmentfolks.
I think portfolio management,business intelligence, reporting
there's a really clearopportunity there.
That's something we areinvesting in.
In terms of where's power, ofmy ability performing now
(30:07):
relative to my peers, what'slikely to happen over the next
three to six months?
I think that's a question everyasset manager asks themselves
right now, and there's everyREIT, every PE fund has an AM
function and they're spendingtheir days kind of mulling over
(30:28):
the data and trying to get aheadof their next competitor, the
new entrant into the market.
Well, ai doesn't sleep, doesn'tneed to eat, and once you've
told it how to think about theworld in terms of you know,
you've given it a prompt and yousaid this is the problem I need
you to solve.
Here's how I want you toapproach solving it.
(30:50):
Then you've armed it withcontextual data, whether that's
the NICMAP vision data set.
Maybe you're giving it accessto your financials, maybe you're
giving it access to some otherinformation you've prepared,
like you would see at aninternal BI dashboard.
Well, now you can just haveyour AI set up alerts and have
(31:12):
it ping you when it sees changesin the data.
Oh, this building's staffingexpense has been increasing over
the last few days.
Looks like overtime budget willmiss the overtime budget this
month.
Is there, you know, kind ofping you.
It's not going to know everyanswer.
It can give you some, you know,based on how you've trained it
(31:36):
or what you've prompted it to do, it could say hey, here's three
things you might want to golook into as possible source of
the problem.
The human is still going tohave to go do that.
So fortunately, john, we stillhave jobs.
Uh, but it's a, it's a greatcompanion, it's a great early
warning system there and it justhelps with the ethos of
data-driven decision making.
(31:57):
Now we're seeing some like we'reseeing you brought up using ai
to maybe prevent falls or kindof monitor how people are, how
they're moving around, likethere is some tech out there now
that's using more machinelearning and video analysis to
do that A company called Safelyyou.
They're increasingly well-knownin the space for that exact use
(32:22):
case.
There's customer success toolsor kind of customer engagement
tools that listen to aconversation between a sales
leader at a community and aprospective resident and make
opinions or recommendations onhow those conversations are
happening.
So there's a lot of reallyinteresting and highly specific
(32:47):
use cases coming out with AI inthis space and it gives me a lot
of excitement and hope forwhere we're going as an industry
.
I will caveat that and say Ialso see AI providers out there
who are trying to be everythingto everyone and I think those
(33:07):
are the ones most likely to failand I think those are the ones
most likely to fail.
Llms specifically.
They have their constraints andthey seem to be most effective
and most valuable when appliedin a very narrow use case, and
so that's the thesis that EnigmaVision is taking with AI.
It's how do we help investmentprofessionals, how do we help
(33:30):
operator professionals withspecific jobs within those
companies?
And again, jobs referring toinvestment analysis, regulatory
compliance, talent acquisition,financial analysis, not trying
to replace the VP of sales orthe VP of investment, because
that's just unrealistic.
(33:50):
We're seeing, and we're we'regetting feedback, that this
approach is is working.
Speaker 2 (33:57):
Awesome.
So when?
When did you guys release yourAI module and you?
When did that come about?
Speaker 1 (34:05):
Yeah, absolutely so.
We had a couple of earlytesters who were using it in
2023.
It was a Definitely a work inprogress at that time, as we
were kind of ironing outeverything.
But January 1st 2024 was ourkind of official release and
then we did a push, kind ofreally started pushing it to the
(34:29):
market at large in March ofthis year.
So it is hot off the presses,but we've got a nice mix of
investors, operators and REITsusing it today.
Speaker 2 (34:39):
Awesome.
Well, I went through one ofyour demonstrations and I
thought it was fantastic whatyou guys have done, so congrats
on really shifting real quicklyto what's needed in the
marketplace and using AI to dothat.
That's awesome, kyle.
Yeah, thank you.
Let's talk about the latestdata trends from NICMAP, vision
(35:00):
Data and what's going on in themarketplace today and where you
see the trends.
Speaker 1 (35:07):
Yeah, absolutely.
So it's fun to talk about thedata side, ai.
I enjoy the products growing.
We're doing a lot of coolthings, but the data is reality.
The data is the truth.
You know it's a measure, it'struth and fortunately, you know
well, my job when it comes tothe data is just tell the story.
You know what do we see andfortunately the story recently
(35:30):
has been a very good one and avery positive one, so it makes
it that much easier and thatmuch more fun to tell.
Last week we just released ourfirst quarter 2024 market
fundamentals data which showedoccupancies improving in our
primary markets, which is the 31largest metros in the US, and
(35:52):
that marks 11 straight quartersof occupancy improvement since
the COVID pandemic, which feelsreally nice to see in the
industry rebounding, to see somecustomers making money and
maybe more importantly than thatis a lot of businesses feel
good, they're performing well.
(36:12):
But the optimism in theindustry of we survived the
pandemic, we've made it throughan incredibly tough series of
years and there's light at theend of the tunnel we're kind of
approaching that piece whereit's like, hey guys, it's not
only at the end of the tunnel,but it's here.
We're enjoying that.
So at a high, high level,occupancies have been rising.
(36:35):
That's being driven by threequarters of just massive
absorption.
When we look at the data, thetrough of COVID, following that
trough, when there weresometimes forced closures at a
state level or some states werepreventing move-ins or new
admissions and so that obviouslyput a huge dent on the
(36:59):
occupancy and the performance ofcommunities.
And following that there wasthis wave of pent-up demand
where we had a few quarters in2021 and 2022 just absolutely
rip on a demand basis as peoplewere moving back in that they
had been forced to stay at homelonger than originally planned.
That COVID rebound slowed downa little bit in early 24, but
(37:26):
starting in third quarter of 23and carrying all the way through
first quarter of 24, we've seenanother big surge in absorption
come back to the market.
Our hypothesis right now isthat is actually the early wave
of the baby boomers coming intothe space, or maybe the last
(37:50):
part of the wave of the silentgeneration coming in and moving
in.
And when we look out the nextfew years, we think that that
demographic wave is going tocontinue and maybe only
accelerate in terms of itsadoption of senior housing.
A couple more specific datapoints I can share with you is
(38:11):
we have seen total occupiedunits so talking about numbers
and not percentages here hitrecord highs in the last few
quarters.
So even though we're not backto pre-pandemic occupancy rates
for every market, a good numberof our primary markets are at or
above their pre-pandemicoccupancy rate.
(38:33):
Our aggregate occupied units orstock is out or near an
all-time high right now and Ithink that's a really important
thing to note because it'stelling us that the industry is
growing.
We're just adding more supply,as we should be, given the
expected demand in the future,but I think it's a promising
(38:56):
sign future, but I think it's apromising sign.
Additionally, we're seeing kindof two different stories between
the primary markets, whichwould be the 31 largest metros
in the US, kind of sorted bypopulation, and what we would
call our secondary markets,which would be the 32nd through
the 99th largest metros in theUS.
(39:17):
Those secondary markets areactually recovering faster than
their primary counterparts,largely driven by what we call
majority assisted livingcommunities.
So those are campuses orcommunities that have assisted
living on site and they couldhave IL or skilled nursing or
memory care on the camps as well.
(39:39):
But if you look at theirprimary unit count.
It's coming from assistedliving that's really driving the
recovery as opposed toindependent living.
So hinting at that needs-baseddriven demand maybe not
surprising, but just interestingto note.
Not surprising, but justinteresting to note.
Additionally, we're seeingconstruction starts kind of
(40:02):
remained near all-time lows on apercentage basis.
Looking at last quarter's dataspecifically so this would be
for fourth quarter 2023,year-over-year inventory in our
primary markets only grew by1.4%, which is near the recorded
(40:22):
lows and the smallest increasewe posted since 2012.
So we're at this kind ofinteresting pivot in the market
right now where the aging boomof, you know, the baby boomers
or the aging wave, as I've seenit be referenced a couple of
times is started.
(40:42):
I suspect it's starting tobreak now, where we're kind of
seeing the early parts of itenter the space.
I think at large, everyoneagrees 2026, 2027 is when the
baby boomer generations maybeofficially entering the senior
housing industry as a potentialconsumer and while that's
(41:07):
literally weeks away, monthsaway, we are delivering an
insufficient amount of supply,of new supply, to meet that
demand.
So we're doing some analysisright now on what the next 10 to
30 years is going to look likefor the industry on a market
fundamentals basis, and one ofthe things our CEO, eric Morton,
(41:32):
has talked about publicly is ifyou just hold penetration rates
constant from 23 to 2050, andyou look at the best year of
construction activity that we'veever delivered as an industry,
as an industry advocate, if wewere to deliver our best year of
(41:53):
construction every year until2050 and our industry
penetration rate was remainingconstant, we would be short
something like $800 billionworth of development between now
and 2050.
So we're still deliveringhundreds of billions of dollars
(42:14):
of assets in that timeline, butwe would be very, very short of
what's expected to be needed bythe time.
The aging wave has kind ofimpacted the industry at large.
So there's a lot of work thatwe need to do collectively and I
think if you're consideringentering the senior housing
(42:36):
market, you should take a goodhard look at it right now,
because we're entering anopportune time for medding.
It's not going to be an easypath.
I don't want to lead you astrayon that, but there's a lot of
upside that we see in this spaceand I'm sure you're seeing the
same thing, john.
Speaker 2 (42:58):
We've had a number of individuals on the podcast.
I've stated the number that youall stated was 775,000 units
are needed by 2030.
And this is six years.
How are we going to do that?
I mean, the current operatorsand owners are happy because
they are getting their occupancyup, because there's no new
(43:20):
product coming up on themarketplace.
So our others that are doingconstruction currently or
constructing larger buildings orsmaller buildings, whatever it
may be how are they doing ittoday?
How are they getting units outof the ground?
Speaker 1 (43:36):
Yeah, it's a good question.
I don't have a firm answer foryou, unfortunately.
It's depending on the marketyou're in.
I'm based in North Carolina.
The development game here is alittle bit different than
California or New York, and ifyou go to Kansas City, if you go
to Minneapolis, if you go toDallas, texas, it's a different
(43:57):
story in each place.
So this business is definitelyone that thrives under kind of
regional focus.
I would say.
I think you know local andregional focus tends to work
best.
Not only do we have this largeneed with very like different
market behaviors across the US,but we're also going to need
(44:19):
more players in order to meetthat need too.
So it'll be interesting to seehow existing constituents
respond, kind of take on thechallenge.
I would expect to see an influxof new market participants over
the next decade as well.
Speaker 2 (44:38):
Yeah, both your firm and our firm are there to
support them, and so we want tomake sure that individuals
listening to this podcast reachout to Kyle and his team, look
into nickmattvisioncom and lookat and see their product set.
It's absolutely the best in theindustry, and so you know.
(45:02):
Let's recap real quickly theimportance of having a trusted
AI partner who understands theindustry.
Speaker 1 (45:11):
Yeah, absolutely so.
If you're going to come andinvest in senior housing,
whether you're going to developa property or you're going to
acquire an existing portfolio,whether you're spending 5
million or 500 million, you haveto have confidence that what
you're buying or building is theright fit, it's in the right
(45:32):
market and that it's going to besuccessful.
And why wouldn't you want thebest data in the industry to
help you make that decision andutilize AI to make your process
go faster and smoother?
Have decades of senior housingoperating, investing,
(45:59):
development and managementexperience.
We have a best-in-classtechnology tool and team.
We have a purpose-builtanalytics system for helping you
with your financial analysis,your site selection, your asset
management and, in some cases,even your sales and marketing
(46:20):
workflows Makes your life easier.
I had a data partner.
I don't want to talk my book myown book too much, but at a
minimum, I recommend asking alot of questions and doing your
diligence, whether you use atool like mine or you use a
partner or, like John, a toollike mine or you use a partner
(46:42):
or, like John, just bethoughtful in your approach.
Speaker 2 (46:43):
Well, it is quite amazing how, number one, you
have the best tool on themarketplace.
Not only that, you're improvingthat product every day and it
really tells how much you allhave a center of excellence
within your company.
So it has really helped us beable to save individuals that
(47:05):
we're consulting with a ton ofmoney Anyone from your
individual that is looking tobuild or develop or acquire a
12-bed all the way up tohundreds and hundreds of beds.
I mean, it's just the data setis just so robust.
So we never debate within Havenwhether or not we're going to
(47:26):
not use NICMAP Vision as oursoftware provider.
There is no debate at all.
It is the best of the best.
So how do people get in touchwith your company or you, or how
do they start the process ofunderstanding how they can
acquire your software?
Speaker 1 (47:44):
The fastest way would be to go to our website at
nickmapvisioncomN-I-C-M-A-P-V-I-S-I-O-Ncom.
If you want to learn about AInickmapvisioncom slash AI.
If you want to learn about AInickmattvisioncom slash AI.
I'm also available on LinkedInand happy to connect there and
talk more.
But we've got a great team ofproduct experts who can sit down
(48:07):
and kind of learn from you whatyou're trying to achieve, where
you're at right now in yourtechnology lifecycle and what
you want to accomplish over thenext six to 24 months.
And you know, if we're a goodfit, we'll make some
recommendations on how to getstarted.
But we'll be transparent withyou too if it's not a good time
and we'll kind of go from there.
Speaker 2 (48:27):
Awesome.
Well, it's been great to haveyou on.
I've wanted to do this for thelast three years.
What a great time for you to beon the show is when AI is
coming out in your product.
So thank you, thank you.
Thank you so much for beingwith me today, kyle, and let's
see where things go over 2024and 2025.
(48:48):
It's going to be a wild ride,so have a great day, kyle, thank
you.
Speaker 1 (48:56):
Thanks, john Cheers.
so when you go and ask the model to do something,
say, hey, chat, gbt, give me arecipe for turkey dinner.
It goes into its memory to saygreat, I've seen thousands or
hundreds of thousands of recipesand let me find ones that I
understand are connected tocooking turkey or something like
(00:28):
that, and I'll give you anoutput.
Well, for senior housing itmight be.
Hey, nai, go and read theregulations for assisted living
in the state of North Carolinaand then tell me what are the
licensing requirements or thestaffing requirements for
medical professionals.
(00:49):
And so it'll go into its memorybanks or go into a reference
document, consume thatinformation and then produce you
an answer.
Speaker 2 (01:04):
Welcome to the Senior Housing Investors Podcast.
If you are an owner operator,investor, developer or buyer of
senior housing, you've come tothe right place.
The best way to stay connectedwith us is to sign up for our
weekly newsletter athavenseniorinvestmentscom.
This podcast doesn't existwithout you, our community.
(01:26):
Thank you for listening andreach out to us anytime.
Speaker 1 (01:36):
Welcome back everyone .
Today, john Haber is speakingwith Kyle Gardner, the Chief
Operating Officer at NICMAPVision.
They have an interestingconversation about data and
decision-making and how AI isplaying a role in that.
Speaker 2 (01:50):
John.
Thanks, Kelsey.
Today, I have the pleasure ofsitting down with Kyle Gardner,
the Chief Operating Officer ofNICMAP Vision.
He was also Chief Operatingofficer of Vision LTC, which was
acquired by Nick in 2021.
His current responsibility isproviding market analysis tools
(02:12):
and comprehensive data toclients in the senior housing
sector.
Welcome to the show, Kyle.
Speaker 1 (02:18):
Yeah, thanks for having me on John Nice to be
with you.
Speaker 2 (02:21):
Yeah, absolutely, man .
You and I go back four yearswhen you were with Vision LTC
and I called out of the blue andsaid my understanding is you
have an amazing tool, and sotell us a little bit about the
origins of both Vision LTC andthen, as you got acquired by
Nick, tell us that story.
Speaker 1 (02:40):
Yeah, absolutely, nick tell us that story?
Yeah, absolutely.
So I joined Vision LTC back in2017, when it was a small but
promising startup and our goalwas to provide kind of market
analytics tools for investorsand operators and developers
focused on the senior housingspace.
(03:01):
On the senior housing space,there was another product in the
market that was much bigger andmore well-known than us, known
as NICMAP, which provided marketfundamentals, inventory rate,
occupancy absorption metrics and, on the Vision LTC side of the
space, what we found is a lot ofour customers were also NICMAP
(03:21):
customers, so it was a reallynice compliment to each other.
You know the ketchup and hotdogs kind of combination, and so
we were growing the Vision LTCbusiness and then had an
opportunity to join forces withNick, and so we created a deal
where Nick purchased Vision LTCand spun us out into a separate
(03:46):
entity that's now known asNickMap Vision, where the
product suite of what wasformerly Vision LTC and formerly
NickMap is all under one roof,and we've spent hundreds or
(04:15):
maybe even thousands of manhours and a lot of money
combining the products into oneplatform psychographics,
inventory rate, occupancymigration patterns of seniors,
healthcare utilization withinsenior housing communities, the
relationship betweenconstruction activity and
(04:35):
changes in market demand and alot more position ourselves to
serve any stakeholder who has aninterest in the real estate of
senior housing, whether you'reinvesting in it, building it or
operating it, or even selling toit In some cases.
We have data to help you andrecently we've started expanding
(05:01):
into AI as a data provider.
It makes sense and we'll talkabout that, I think, here in a
few minutes.
But we've also got some toolsaround what I call a value add
for our customers.
We have a listings platformwhere customers can buy, sell
and trade senior housing assets.
That's no cost to them, no costto the broker, no cost to the
(05:23):
seller or the buyer.
You just have to be a client ofthe platform to get access.
We've got some healthcare datato help operators become aware
of who the healthcare providersare in the space to artificial
intelligence, specificallythrough OpenAI's GPT models, to
(05:47):
bring AI-powered automations tosenior housing investors and
operators.
So we have a very wide productfootprint, but we focus solely
on the senior housing space andour customers have proven with
their wallets and their feedbackthat we're on the right track.
Speaker 2 (06:05):
So Well, I can say that when we started with Vision
LTC as a company at HavenSenior Investments, we were just
blown away.
And then doing the combinationof both Nick and Vision's data
sets combined has given us theability to really advise our
(06:27):
clients wisely with data andalso on the brokerage side of
our business, being able tounderstand the markets all the
way down to the smallest marketsin the United States, even the
small towns.
It is a beautiful package thatyou all have, and I'm very
excited to learn more about theAI side of the business.
(06:49):
And so you know how does AIplay a role in allowing
operators and investors toaccess data for decision-making,
and I'd love to understand whatthe challenges are currently
and how this AI addresses thosechallenges.
Speaker 1 (07:07):
Yeah, yeah, absolutely.
So to answer that question, letme start first with kind of the
challenges at large that ourdata and our tools solve,
because AI is really anextension of that and a
continuation of that Mission'sthe wrong word, but kind of the
service that we play.
(07:27):
So if you're a real estateinvestor whether it be a
first-time fund entering seniorhousing or you're a REIT
purpose-built for this industryyou need information in order to
make investment decisions right.
So 20 years ago the way to getthat information was to hire a
(07:49):
consultant or a broker to go andput boots on the ground and
maybe live in the market for afew days and do some recon.
You'd probably spend some timethere as well, look at Census
Bureau data and so on, and thenthey'd bring you a large report
and you'd kind of debrief withthem over time and kind of
fast-forwarding to now.
The capability of analyticstools and demography and other
(08:14):
analytics sets like KnickknackVision and like GIS technologies
have enabled kind of desktopstudies where you don't have to
leave the office to get at leastlike a fingertip feel of the
population size and what's inthe market, and Google Maps
gives you some pile of visuals.
You're still putting boots onthe ground before you write the
(08:36):
check.
But the time it takes to accessthat information has gone from
maybe weeks and months to days,and now with AI and the amount
of data that we have in our tool, it can go down to minutes.
And so the challenges thatwe're helping solve is really
helping investors make informeddecisions with more confidence
(08:58):
and in less time, like that'sthe motto that we're kind of
telling ourselves here atNickmap Visions.
That's the motto that we'rekind of telling ourselves here
at Nickmap Visions how do wehelp people execute faster to
either get to an informed no andwalk away from a bad deal, get
to an informed yes and takeadvantage of an opportunity for
their competitors and maybe evenhelp them define what a good
(09:19):
deal or a bad deal is for them.
That's going to changedepending on the shop and risk
preferences and things of thatnature.
But information and intelligenceis kind of at the core of all
we do.
We've been servicing customersfor 20 years on that theory,
going back to the early days ofNICMAP, and now that AI is here,
it just allows us to acceleratethat process even faster, but
(09:44):
it also lets us go considerablydeeper into detail, right?
So I think everyone's familiarwith Census Bureau data or
Bureau of Labor Statistics data.
You know tables on tables ofnumbers and you know CSV
spreadsheets nothing really sexy, but it's the source of truth
that we have.
People are probably familiarwith like Claritas and Esri
(10:07):
those are two data partners thatwe use in our systems and that
data is available.
It's available very quickly,especially in systems like
NICMAP Vision, but it requiresyou as the analyst to make an
opinion.
You know, read the data andthen figure out is this good,
bad, is this normal, is it anoutlier?
(10:30):
Right?
With AI combined with kind ofindustry experience, we can help
the customer go frominformation to opinion or go
from information torecommendation very quickly, and
so AI is kind of an extensionof the you know data as a
(10:52):
service.
Our CEO likes to use the phraseyou know, ai is the arms and
legs and the data is the body.
You know, so it helps make itmakes it more actionable Awesome
makes it more actionableAwesome.
Speaker 2 (11:11):
And to tell our audience a little bit about just
overall artificial intelligenceI'm sure many know what it
means or what it is or just givea high level overview of
artificial intelligence usingOpenAI, chatgpt 4, and here
shortly ChatGPT5 is coming out.
So pretty cool stuff.
So if you could describe kindof in a layman's term what you
(11:35):
mean by AI, yeah, absolutely so.
Speaker 1 (11:39):
Well, sam Altman, ceo of OpenAI, if you're listening,
please release GPT-5 soonerthan later.
But what is AI?
What is a large language model?
I think the average person'sheard about ChatGPT, at least in
passing, since it was releaseda few years ago.
At its core, ai, or largelanguage AI, is a type of
(12:04):
technology.
Under that umbrella or underthat segment of technology,
there's a couple of differentapplications.
There's large language models,machine learning, natural
language processing and a fewother technology applications,
and I am not on the technicalside of the business so I won't
(12:26):
try to explain the nuances ofall of those.
But the way that I understandit, the way that I bring to my
clients and working with our AIengineering team they're the
true geniuses here is largelanguage models is basically a
statistical model that you'vetrained an application to read
(12:48):
and consume information.
So whether that's a pressrelease, a blog article, the
content on a webpage, if youwere to take a picture of a menu
you know at a restaurant andyou could convert that picture
to text, you know the languagemodel could read that text and
consume it.
And then there's kind of amethod where the technology is
(13:20):
reading the text that's on thescreen or on the page and
putting looking at, you know,open AIs GPT-4 is currently, you
know, in April of 24, whenwe're recording this is
currently considered kind of themarket leader in model
performance.
There's other models fromAnthropic and Meta and Google,
(13:41):
but what they're all doing isthey're being trained on a
certain type of data set andthen those models take what
they've learned and, much like ahuman, try to apply that to
complete certain practices orcomplete certain outcomes or
create certain outcomes.
And I know this is probably alittle confusing here, so let me
(14:03):
tie this back together tosomething that real-world
example that makes sense.
So OpenAI goes and trains thesemodels on billions of data
points webpages, pictures,videos.
Just think about the content onthe internet, and OpenAI has
probably seen it or touched itin some capacity newspapers,
(14:24):
blogs, so on and so forth.
As it's reading all thiscontent, it's gathering
information in terms ofknowledge or facts, but it's
also gathering information interms of how the human language
is transcribed and written.
So when you go and ask themodel to do something, say
(14:47):
written.
So when you go and ask themodel to do something, say, hey,
chatgpt, give me a recipe for aturkey dinner.
It goes into its memory to saygreat, I've seen thousands or
hundreds of thousands of recipes, and let me find ones that I
understand are connected tocooking turkey or something like
that, and I'll give you anoutput.
Well, for senior housing andapproach you know where it
(15:08):
becomes applicable for investorsand operators, it might be.
Hey, nai, go and read theregulations for assisted living
in the state of North Carolinaand then tell me what are the
licensing requirements or thestaffing requirements for
medical professionals in thestate, what's the ratio I need
(15:31):
of a med tech or an RN or an LPNto certain residents or any of
those positions even require, soon and so forth?
And so it'll go into its memorybanks or go into a reference
document, consume thatinformation and then produce you
an answer.
It's incredibly powerful, it'sincredibly complex, but at the
(15:51):
end of the day, it's a type oftechnology, albeit it's a
cutting edge one.
So it feels new, it feels, youknow, in some cases kind of
scary, but it's code at the endof the day.
Speaker 2 (16:06):
Well, it's interesting that you spoke about
.
You know, tell ChatGPT for aquestion and it'll go into its
memory and bring out output.
Well, I did that before theshow.
I wanted to quote on whatChatGPT for would come up with
in regards to AI itself, andthey came up with an optimistic
(16:27):
vision, ethical consideration,transformational change but I
like this one the best, and it'sa quote from Stephen Hawkins,
and AI will be the best or worstthing ever for humanity, so
let's make it the best.
And it's so true.
It's really the individualsthat are training these AI
models.
Are you going to train it to bethe best it can be and be the
(16:50):
best for humanity, or are yougoing to create it in a fashion
that is detrimental to us?
I believe in many other beliefsthat it's one of the most
transformational changes thatwe're going to encounter in the
history of the world, and solet's get back to where AI can
(17:11):
add value to the senior housingecosystem today.
Can you riff on that a littlebit?
Speaker 1 (17:18):
Yeah, absolutely.
If you look at the jobs and Idon't mean like job titles or
the positions that people have,I literally mean like the work
to be done by people who workfor a private equity company or
an investment manager or anoperators management company
(17:42):
Look at the job content thatthey're doing every day.
Yeah, if you had to boil itdown to really simple concepts,
are they reading, writing, doingmath?
A lot of work that is done inthe investment workflow and the
asset management workflow comesdown to reading something,
(18:07):
reading it once and committingit to memory, or coming back to
it multiple times, applyingcritical thinking and then
creating an action plan andgoing from there.
Or it's read something, go andbuild a model in Excel to
(18:27):
represent the world and do somescenario analysis of what could
or might happen, and then createan action plan from there, and
that's a very oversimplified wayof thinking about the world.
But when you look at the jobsto be done or the work to be
done by professionals in thisindustry and you realize we
spend a lot of time readingmanagement contracts,
(18:48):
regulations, laws.
We spend a lot of time lookingat offering them randoms, broker
opinions of value, pitch booksand pitch decks and things of
that nature, and then multipleparties are asked to make in
most cases million dollars,sometimes tens of million
dollars decisions on thatinformation.
(19:12):
But what is AI really reallygood at?
It is exceptional at reading,consuming information and
applying a kind of a criticalthinking to that information,
with some caveats.
There has been research overthe last year that, like Chat,
gbt or gbt4 those are the samemodels, by the way that gbt4 can
(19:37):
pass the bar exam thatpracticing attorneys have to
take in the us at a level higherthan the average lawyer.
Yeah right, because it's your,it's reading and critical
thinking and and analysis, butit's really bad at math.
Llms are terrible at math.
So it's your, it's reading andcritical thinking and and
analysis, but it's really bad atmath.
(19:58):
Llms are terrible at math, soit's kind of funny that it can't
do third grade math properly,but it can pass the bar exam,
which is, in certain circles,right Considered kind of one of
the highest white collar jobs ormost prestigiouscollar jobs
that we have in the US.
And so if we play to AIstrengths, it can help us
accelerate our decision-makingtime in certain areas.
(20:22):
It can help us dig deeper intosubject knowledge.
It can help us do scenarioanalysis in many different ways,
and if I was going to breakthose out into a couple of
different examples, let's lookat an investment analyst or a VP
of investment at a firm.
They're reading OMs, they'rereading contracts, they're
(20:46):
drafting OMs, they're reviewingBOVs all to make a decision.
Well, with NICMAT Vision, we'vetaken AI and our deep industry
expertise and createdautomations that use a mixture
of LLMs, like an open AI, not achat GBT or, sorry, a GBT4.
(21:07):
Combine that with our kind ofown proprietary technology is
where, like our engineering team, our developers have created
code-based workflows that arewhat I'll call rigid, meaning
that they're not llm based, thatone plus one is always two, uh.
(21:28):
So we've combined kind of thepower of that custom technology
with the creativity of the LLMsto focus on specific automations
.
And so one of the automationsthat our investors have been
using and love is take aoffering memorandum, just upload
(21:48):
it in any readable format sothe PDF has to be readable or it
can be a Word document.
We'll use the LLM to read theoffering memorandum, pull out
the information that'sinteresting and relevant, based
on some customizable promptsthat our team's created, and
(22:09):
then draft kind of a two page,three page executive summary and
as part of that summary thatthe automation's creating, we're
injecting NICMAP vision datainto it.
So your broker sends you an OM,you load it into the tool.
You answer a couple of basicquestions what's the name of the
(22:31):
subject property?
What do you want to ask theoffering memorandum?
The same way, you kind of entera prompt in the chat GPT, and
then a couple of other basicparameters.
I think there's maybe fivequestions total.
You hit run and in five to 10minutes you know the time it
takes to go refill your coffee.
(22:52):
You come back and you've got athree-page executive summary of
the offering memorandum.
Looking at the subject property, it's going to analyze any of
the financial information that'sin there.
It's going to add market compsfrom our NICBAT Vision market
fundamentals data set.
It's going to pull indemographics specific to your
market area.
It'll even give you questionsthat you should send back to the
(23:14):
broker or to the seller so youcan dig deeper into the product
or, sorry, into the investmentopportunity and kind of go from
there.
Is that automation going toreplace your investment team?
No, absolutely, absolutely not.
It's going to replace yourinvestment team.
(23:36):
No, absolutely, absolutely not.
It's going to make them a wholehell of a lot faster at what
they do, though, because now,instead of waiting, a
well-staffed analyst team who'sputting in long hours maybe
charge that same analysis aroundin two business days.
Maybe say, an average teammaybe takes a whole week.
I'm talking about like a true,like deep dive into the, the OM,
(24:00):
the underlying data, uh, doingtheir own third party analysis.
You know they maybe takes thema week, and you're getting that
same kind of V1 draft in 10minutes.
So imagine what your team could.
You know that same team who ispumping out reports in two to
five days.
You've just given them thatwhole.
(24:21):
You know, let's call it 48hours back minimum, to then go
even deeper or to go look atmore deals, and what could that
do for your, your process?
What could that do for yourprocess?
What could that do for yourcompetitiveness?
I think it can do a lot.
There's other things we're doingas well, just at a high level.
(24:43):
We've created automations tohelp with reviewing income
statements and readingfinancials, providing summary
analyses of those, basicallyhelping you stay ahead of market
trends.
We've got some tools built onthe regulations, obviously
senior housings, regulated atthe state level by a large,
(25:04):
specifically for assisted livingand memory care.
So the regional and nationalproviders have a lot of rule
books that they have to keephandy and, fortunately, ai is
quite good at analyzing those,and we've created some workflows
that, if every building isgetting surveyed and if you're
(25:26):
getting deficiencies, back atsome point, you're going to get
asked to create a plan ofcorrections.
Back at some point, you'regoing to get asked to create a
plan of corrections.
Well, why not have AI do yourfirst draft and then have your
legal experts spend their timevetting the information and
updating the workflow to make itabsolute and applicable to the
(25:46):
reuse case?
But instead of spending theirtime drafting what is basically
a template form, let's put AI towork there.
Or, if you've got questions ofthe regulations, instead of
spending hours and hours goingthrough these terrible websites,
the hundreds of thousands ofclick-through links that the
(26:08):
states love to utilize, why notjust ask a question of a chat
bot and get an answer back inless than a minute?
It's quite powerful.
So everything we're doing withAI really ties back to
accelerating yourdecision-making process and
giving you quick, strategicmarket insights.
Ai is not going to replace yourjob today, and if someone's
(26:33):
telling you that they have notspent enough time using AI, but
if your team is using AI, youwill outperform a team who is
not plain and simple.
Speaker 2 (26:44):
Agreed Bottom line is I always felt that AI was going
to be a companion to us ashuman beings, and that just
makes everything more powerfulour ability to how much they're
walking or how much they aren'twalking.
Maybe that decision makingthrough AI can alert us in the
(27:19):
future to areas that we have noidea we would have uncovered if
we didn't have AI next to us.
So tell us about the future ofthe senior housing ecosystem
when it comes to AI.
What's the future thinkingthat's going on behind the
scenes at NICMAT Vision?
Speaker 1 (27:38):
Yeah, absolutely.
I'll give you a two-part answeron this.
I'll talk a little bit aboutwhat we're working on and how
we're thinking about things.
But I feel like I'd be doingthe pot of disservice, john, if
I didn't share a little bitabout what I'm already seeing in
the market from other providersin the AI space and kind of
(27:58):
where the industry is going atlarge.
I feel kind of unique andprivileged to get to see quite a
lot from the different NICconferences and stuff that I've
been to and it's quiteimpressive and stuff that I've
been to and it's quiteimpressive.
So for NICMEM Vision, we see AIas being a I like your word
(28:19):
like a partner.
We see it as being an enablerfor team members to accelerate
their workflow.
The most value to be had rightnow in 2024 is for research and
customer service and labormanagement, asset management,
investment management workflows.
(28:40):
So whether that might besomeone in the FP&A department,
that might be someone on theacquisitions team at a REIT, it
might be the asset manager of aninvestment fund, it could be
the VP of HR, vp of people at anoperator or something of that
nature.
Those are the areas that we seeit kind of living in right now
(29:02):
and it kind of goes back to whatI was talking about before with
.
It's very good at reading andcritical thinking and doing some
scenario analysis.
And doing some scenarioanalysis when it's going, is it
kind of leading us to think thatthere's potential to help with,
maybe, labor management orlabor assistance at the
community level.
At the management level, thatwe're probably going to see that
(29:26):
in the shape of staffingoptimization.
We'll probably see that interms of embedded features in
current applications, whetheryou're EHR, erp.
I think there's some very clearopportunity there.
The ecosystem at large that'snot something NCHMAP Vision's
(29:47):
likely to want to touch on, butwe see that as kind of an
opportunity for the investmentfolks.
I think portfolio management,business intelligence, reporting
there's a really clearopportunity there.
That's something we areinvesting in.
In terms of where's power, ofmy ability performing now
(30:07):
relative to my peers, what'slikely to happen over the next
three to six months?
I think that's a question everyasset manager asks themselves
right now, and there's everyREIT, every PE fund has an AM
function and they're spendingtheir days kind of mulling over
(30:28):
the data and trying to get aheadof their next competitor, the
new entrant into the market.
Well, ai doesn't sleep, doesn'tneed to eat, and once you've
told it how to think about theworld in terms of you know,
you've given it a prompt and yousaid this is the problem I need
you to solve.
Here's how I want you toapproach solving it.
(30:50):
Then you've armed it withcontextual data, whether that's
the NICMAP vision data set.
Maybe you're giving it accessto your financials, maybe you're
giving it access to some otherinformation you've prepared,
like you would see at aninternal BI dashboard.
Well, now you can just haveyour AI set up alerts and have
(31:12):
it ping you when it sees changesin the data.
Oh, this building's staffingexpense has been increasing over
the last few days.
Looks like overtime budget willmiss the overtime budget this
month.
Is there, you know, kind ofping you.
It's not going to know everyanswer.
It can give you some, you know,based on how you've trained it
(31:36):
or what you've prompted it to do, it could say hey, here's three
things you might want to golook into as possible source of
the problem.
The human is still going tohave to go do that.
So fortunately, john, we stillhave jobs.
Uh, but it's a, it's a greatcompanion, it's a great early
warning system there and it justhelps with the ethos of
data-driven decision making.
(31:57):
Now we're seeing some like we'reseeing you brought up using ai
to maybe prevent falls or kindof monitor how people are, how
they're moving around, likethere is some tech out there now
that's using more machinelearning and video analysis to
do that A company called Safelyyou.
They're increasingly well-knownin the space for that exact use
(32:22):
case.
There's customer success toolsor kind of customer engagement
tools that listen to aconversation between a sales
leader at a community and aprospective resident and make
opinions or recommendations onhow those conversations are
happening.
So there's a lot of reallyinteresting and highly specific
(32:47):
use cases coming out with AI inthis space and it gives me a lot
of excitement and hope forwhere we're going as an industry
.
I will caveat that and say Ialso see AI providers out there
who are trying to be everythingto everyone and I think those
(33:07):
are the ones most likely to failand I think those are the ones
most likely to fail.
Llms specifically.
They have their constraints andthey seem to be most effective
and most valuable when appliedin a very narrow use case, and
so that's the thesis that EnigmaVision is taking with AI.
It's how do we help investmentprofessionals, how do we help
(33:30):
operator professionals withspecific jobs within those
companies?
And again, jobs referring toinvestment analysis, regulatory
compliance, talent acquisition,financial analysis, not trying
to replace the VP of sales orthe VP of investment, because
that's just unrealistic.
(33:50):
We're seeing, and we're we'regetting feedback, that this
approach is is working.
Speaker 2 (33:57):
Awesome.
So when?
When did you guys release yourAI module and you?
When did that come about?
Speaker 1 (34:05):
Yeah, absolutely so.
We had a couple of earlytesters who were using it in
2023.
It was a Definitely a work inprogress at that time, as we
were kind of ironing outeverything.
But January 1st 2024 was ourkind of official release and
then we did a push, kind ofreally started pushing it to the
(34:29):
market at large in March ofthis year.
So it is hot off the presses,but we've got a nice mix of
investors, operators and REITsusing it today.
Speaker 2 (34:39):
Awesome.
Well, I went through one ofyour demonstrations and I
thought it was fantastic whatyou guys have done, so congrats
on really shifting real quicklyto what's needed in the
marketplace and using AI to dothat.
That's awesome, kyle.
Yeah, thank you.
Let's talk about the latestdata trends from NICMAP, vision
(35:00):
Data and what's going on in themarketplace today and where you
see the trends.
Speaker 1 (35:07):
Yeah, absolutely.
So it's fun to talk about thedata side, ai.
I enjoy the products growing.
We're doing a lot of coolthings, but the data is reality.
The data is the truth.
You know it's a measure, it'struth and fortunately, you know
well, my job when it comes tothe data is just tell the story.
You know what do we see andfortunately the story recently
(35:30):
has been a very good one and avery positive one, so it makes
it that much easier and thatmuch more fun to tell.
Last week we just released ourfirst quarter 2024 market
fundamentals data which showedoccupancies improving in our
primary markets, which is the 31largest metros in the US, and
(35:52):
that marks 11 straight quartersof occupancy improvement since
the COVID pandemic, which feelsreally nice to see in the
industry rebounding, to see somecustomers making money and
maybe more importantly than thatis a lot of businesses feel
good, they're performing well.
(36:12):
But the optimism in theindustry of we survived the
pandemic, we've made it throughan incredibly tough series of
years and there's light at theend of the tunnel we're kind of
approaching that piece whereit's like, hey guys, it's not
only at the end of the tunnel,but it's here.
We're enjoying that.
So at a high, high level,occupancies have been rising.
(36:35):
That's being driven by threequarters of just massive
absorption.
When we look at the data, thetrough of COVID, following that
trough, when there weresometimes forced closures at a
state level or some states werepreventing move-ins or new
admissions and so that obviouslyput a huge dent on the
(36:59):
occupancy and the performance ofcommunities.
And following that there wasthis wave of pent-up demand
where we had a few quarters in2021 and 2022 just absolutely
rip on a demand basis as peoplewere moving back in that they
had been forced to stay at homelonger than originally planned.
That COVID rebound slowed downa little bit in early 24, but
(37:26):
starting in third quarter of 23and carrying all the way through
first quarter of 24, we've seenanother big surge in absorption
come back to the market.
Our hypothesis right now isthat is actually the early wave
of the baby boomers coming intothe space, or maybe the last
(37:50):
part of the wave of the silentgeneration coming in and moving
in.
And when we look out the nextfew years, we think that that
demographic wave is going tocontinue and maybe only
accelerate in terms of itsadoption of senior housing.
A couple more specific datapoints I can share with you is
(38:11):
we have seen total occupiedunits so talking about numbers
and not percentages here hitrecord highs in the last few
quarters.
So even though we're not backto pre-pandemic occupancy rates
for every market, a good numberof our primary markets are at or
above their pre-pandemicoccupancy rate.
(38:33):
Our aggregate occupied units orstock is out or near an
all-time high right now and Ithink that's a really important
thing to note because it'stelling us that the industry is
growing.
We're just adding more supply,as we should be, given the
expected demand in the future,but I think it's a promising
(38:56):
sign future, but I think it's apromising sign.
Additionally, we're seeing kindof two different stories between
the primary markets, whichwould be the 31 largest metros
in the US, kind of sorted bypopulation, and what we would
call our secondary markets,which would be the 32nd through
the 99th largest metros in theUS.
(39:17):
Those secondary markets areactually recovering faster than
their primary counterparts,largely driven by what we call
majority assisted livingcommunities.
So those are campuses orcommunities that have assisted
living on site and they couldhave IL or skilled nursing or
memory care on the camps as well.
(39:39):
But if you look at theirprimary unit count.
It's coming from assistedliving that's really driving the
recovery as opposed toindependent living.
So hinting at that needs-baseddriven demand maybe not
surprising, but just interestingto note.
Not surprising, but justinteresting to note.
Additionally, we're seeingconstruction starts kind of
(40:02):
remained near all-time lows on apercentage basis.
Looking at last quarter's dataspecifically so this would be
for fourth quarter 2023,year-over-year inventory in our
primary markets only grew by1.4%, which is near the recorded
(40:22):
lows and the smallest increasewe posted since 2012.
So we're at this kind ofinteresting pivot in the market
right now where the aging boomof, you know, the baby boomers
or the aging wave, as I've seenit be referenced a couple of
times is started.
(40:42):
I suspect it's starting tobreak now, where we're kind of
seeing the early parts of itenter the space.
I think at large, everyoneagrees 2026, 2027 is when the
baby boomer generations maybeofficially entering the senior
housing industry as a potentialconsumer and while that's
(41:07):
literally weeks away, monthsaway, we are delivering an
insufficient amount of supply,of new supply, to meet that
demand.
So we're doing some analysisright now on what the next 10 to
30 years is going to look likefor the industry on a market
fundamentals basis, and one ofthe things our CEO, eric Morton,
(41:32):
has talked about publicly is ifyou just hold penetration rates
constant from 23 to 2050, andyou look at the best year of
construction activity that we'veever delivered as an industry,
as an industry advocate, if wewere to deliver our best year of
(41:53):
construction every year until2050 and our industry
penetration rate was remainingconstant, we would be short
something like $800 billionworth of development between now
and 2050.
So we're still deliveringhundreds of billions of dollars
(42:14):
of assets in that timeline, butwe would be very, very short of
what's expected to be needed bythe time.
The aging wave has kind ofimpacted the industry at large.
So there's a lot of work thatwe need to do collectively and I
think if you're consideringentering the senior housing
(42:36):
market, you should take a goodhard look at it right now,
because we're entering anopportune time for medding.
It's not going to be an easypath.
I don't want to lead you astrayon that, but there's a lot of
upside that we see in this spaceand I'm sure you're seeing the
same thing, john.
Speaker 2 (42:58):
We've had a number of individuals on the podcast.
I've stated the number that youall stated was 775,000 units
are needed by 2030.
And this is six years.
How are we going to do that?
I mean, the current operatorsand owners are happy because
they are getting their occupancyup, because there's no new
(43:20):
product coming up on themarketplace.
So our others that are doingconstruction currently or
constructing larger buildings orsmaller buildings, whatever it
may be how are they doing ittoday?
How are they getting units outof the ground?
Speaker 1 (43:36):
Yeah, it's a good question.
I don't have a firm answer foryou, unfortunately.
It's depending on the marketyou're in.
I'm based in North Carolina.
The development game here is alittle bit different than
California or New York, and ifyou go to Kansas City, if you go
to Minneapolis, if you go toDallas, texas, it's a different
(43:57):
story in each place.
So this business is definitelyone that thrives under kind of
regional focus.
I would say.
I think you know local andregional focus tends to work
best.
Not only do we have this largeneed with very like different
market behaviors across the US,but we're also going to need
(44:19):
more players in order to meetthat need too.
So it'll be interesting to seehow existing constituents
respond, kind of take on thechallenge.
I would expect to see an influxof new market participants over
the next decade as well.
Speaker 2 (44:38):
Yeah, both your firm and our firm are there to
support them, and so we want tomake sure that individuals
listening to this podcast reachout to Kyle and his team, look
into nickmattvisioncom and lookat and see their product set.
It's absolutely the best in theindustry, and so you know.
(45:02):
Let's recap real quickly theimportance of having a trusted
AI partner who understands theindustry.
Speaker 1 (45:11):
Yeah, absolutely so.
If you're going to come andinvest in senior housing,
whether you're going to developa property or you're going to
acquire an existing portfolio,whether you're spending 5
million or 500 million, you haveto have confidence that what
you're buying or building is theright fit, it's in the right
(45:32):
market and that it's going to besuccessful.
And why wouldn't you want thebest data in the industry to
help you make that decision andutilize AI to make your process
go faster and smoother?
Have decades of senior housingoperating, investing,
(45:59):
development and managementexperience.
We have a best-in-classtechnology tool and team.
We have a purpose-builtanalytics system for helping you
with your financial analysis,your site selection, your asset
management and, in some cases,even your sales and marketing
(46:20):
workflows Makes your life easier.
I had a data partner.
I don't want to talk my book myown book too much, but at a
minimum, I recommend asking alot of questions and doing your
diligence, whether you use atool like mine or you use a
partner or, like John, a toollike mine or you use a partner
(46:42):
or, like John, just bethoughtful in your approach.
Speaker 2 (46:43):
Well, it is quite amazing how, number one, you
have the best tool on themarketplace.
Not only that, you're improvingthat product every day and it
really tells how much you allhave a center of excellence
within your company.
So it has really helped us beable to save individuals that
(47:05):
we're consulting with a ton ofmoney Anyone from your
individual that is looking tobuild or develop or acquire a
12-bed all the way up tohundreds and hundreds of beds.
I mean, it's just the data setis just so robust.
So we never debate within Havenwhether or not we're going to
(47:26):
not use NICMAP Vision as oursoftware provider.
There is no debate at all.
It is the best of the best.
So how do people get in touchwith your company or you, or how
do they start the process ofunderstanding how they can
acquire your software?
Speaker 1 (47:44):
The fastest way would be to go to our website at
nickmapvisioncomN-I-C-M-A-P-V-I-S-I-O-Ncom.
If you want to learn about AInickmapvisioncom slash AI.
If you want to learn about AInickmattvisioncom slash AI.
I'm also available on LinkedInand happy to connect there and
talk more.
But we've got a great team ofproduct experts who can sit down
(48:07):
and kind of learn from you whatyou're trying to achieve, where
you're at right now in yourtechnology lifecycle and what
you want to accomplish over thenext six to 24 months.
And you know, if we're a goodfit, we'll make some
recommendations on how to getstarted.
But we'll be transparent withyou too if it's not a good time
and we'll kind of go from there.
Speaker 2 (48:27):
Awesome.
Well, it's been great to haveyou on.
I've wanted to do this for thelast three years.
What a great time for you to beon the show is when AI is
coming out in your product.
So thank you, thank you.
Thank you so much for beingwith me today, kyle, and let's
see where things go over 2024and 2025.
(48:48):
It's going to be a wild ride,so have a great day, kyle, thank
you.
Speaker 1 (48:56):
Thanks, john Cheers.
Popular Podcasts
Bookmarked by Reese's Book Club
Welcome to Bookmarked by Reese’s Book Club — the podcast where great stories, bold women, and irresistible conversations collide! Hosted by award-winning journalist Danielle Robay, each week new episodes balance thoughtful literary insight with the fervor of buzzy book trends, pop culture and more. Bookmarked brings together celebrities, tastemakers, influencers and authors from Reese's Book Club and beyond to share stories that transcend the page. Pull up a chair. You’re not just listening — you’re part of the conversation.
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!
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