Latest / Women Disrupting Tech / How Cogensus Uses AI to Tackle Loneliness and Cognitive Decline with Raju Joshi | Ep. 112

Transcript

How Cogensus Uses AI to Tackle Loneliness and Cognitive Decline with Raju Joshi | Ep. 112 · Women Disrupting Tech

0:00
Yeah. So I think AI in general has
0:03
great potential if it's developed and applied correctly.
0:08
I think that applications like cogensis that are geared to
0:15
improve loneliness and or social connection have an opportunity
0:21
to provide those insights around a myriad of conditions, not just
0:25
cognitive decline. And so there is an opportunity
0:29
to improve care management for patients to help clinicians with
0:38
data and insights that will hopefully reduce their burden
0:43
and burnout. Welcome to Women Disrupting
0:57
Tech. My name is dear Kanupkas, and on
0:59
this podcast, I give the women who work in tech the platform
1:03
they need to inspire and empower each other.
1:07
Join me on a journey into the world of entrepreneurship,
1:10
funding biases, and role models to discover the strategies for
1:14
success in the tech industry. In other words, how women can
1:19
disrupt tech. Are you ready to listen?
1:28
Hey, welcome back to Women Disrupting Tech and thank you
1:32
for tuning into episode 112. Did you know that loneliness
1:37
increases your risk of serious health problems by 59%?
1:43
Loneliness is one of the most overlooked drivers as we age,
1:47
and women are once again hit the hardest.
1:52
They live alone more often, they provide the majority of
1:56
caregiving hours, and they tend to delay care for themselves.
2:01
That's where my guest today comes in.
2:03
Her name is Raju Joshi and she's the Co founder and chief
2:07
clinical officer of Cogensis. Cogensis is a health tech
2:12
startup that uses AI to support people experiencing loneliness
2:17
and early cognitive decline as well as their caregivers.
2:21
Raju notes this problem personally, she lost 2 senior
2:26
family members in the same year, and even though she has 30 years
2:31
of experience in healthcare, she saw how I'll equipped even the
2:36
best prepared families can be. In this episode of Women
2:40
Disrupting Tech, Rajo and I talk about why loneliness hits women
2:45
harder and what long term effects on health are, how
2:50
Cogenesis uses AI to listen for, what caregivers often miss, what
2:56
ethical, inclusive AI in health looks like in practice, and why
3:01
building trust with users starts with usability, not with
3:06
features. It's a rich conversation about
3:10
loneliness, cognitive decline, technology, and much, much more.
3:14
And my biggest take away was how to have a better conversation
3:19
with my mother. The cue, I don't want to burden
3:22
you. So when you're ready to listen
3:26
and learn how Covjansis uses AI to tackle loneliness and
3:31
cognitive decline, please. Welcome to women disrupting tech
3:35
Raju Joshi. Raju Joshi, welcome to women
3:38
disrupting tech. Great to have you a question I
3:41
always start with. Is how you got into your current
3:44
position in your case as Chief Clinical Officer of Covjansis.
3:49
Thank you so much for having me Durkian.
3:51
I'm very excited to be on your show.
3:55
Yes, I actually joined the cogensis team over a little over
4:00
a year ago. I have a background in
4:02
biomedical sciences and epidemiology and I have been in
4:05
healthcare for over 30 years. I've had the good fortune of
4:09
working in a variety of different sub sectors within
4:13
healthcare including academic, research, payers, long term care
4:19
facility, supporting them. From a pharmacy standpoint, I've
4:26
worked at Medtronic, at Amgen, at a start up with Deloitte.
4:30
I was with Deloitte Consulting for many years and this gave me
4:34
the opportunity to really see healthcare at a very broad level
4:40
as well as in depth within each one of those sub sectors.
4:46
And so I was introduced to Cogensis through a mutual friend
4:52
who indicated that they were looking for somebody with a, a
4:57
clinical background from that standpoint, I mean, having
5:02
knowledge of how healthcare systems work and also understand
5:06
the biomedical sciences behind it.
5:09
So I was advising the company and then through that process
5:13
and being a board member, they asked me to join them in a
5:18
larger capacity as chief Clinical officer to help them
5:23
build a a company and a platform that maintains scientific
5:27
fidelity and integrity and generates the necessary evidence
5:31
to support the products build out and work with clinicians
5:35
within the community to ensure clinical relevance and utility.
5:40
OK, cool. And what what made Gogensis
5:44
interesting to you personally? Yeah, so last year I lost 2
5:50
senior members of my family. My mother-in-law who is 4 foot
5:57
11 and an ICU nurse was very, very bright and very brave.
6:04
She suffered from dementia and I watched her her decline
6:11
initially somewhat slowly and I thought this was just
6:15
age-related kind of, you know, cognitive changes to a, a rapid
6:22
decline. And then ultimately we did lose
6:25
her, as I mentioned in February, and then my father somewhat
6:30
unexpectedly on Father's Day. And he was a pharmacist, A
6:34
pharmacologist. He, he very well understood what
6:36
was happening to him as well. And, and even though I've been
6:40
in this space for 30 years, I have family members who are
6:43
physicians. I have family members who are
6:47
attorneys and understand the legal process as it relates to
6:51
healthcare decisions and healthcare policies.
6:54
We found ourselves I'll equipped to handle their late stage
7:01
conditions. And by that I mean we felt that
7:07
most of the conversations were centered around what medications
7:13
were they taking? What was their mood like in the
7:16
morning? Could we speak with them?
7:18
Often times they were too tired to talk towards those these
7:21
later, you know, months. And we just felt this sense of
7:26
of loss. And so I really wanted to help
7:30
build cogensis in a way that helped to not only meet patients
7:37
where they were people, where they were in their journey, but
7:41
also help preserve that journey for those that they left behind.
7:46
And so legacy preservation is, is one of our unique features.
7:51
And I'll, you know, given an opportunity to go into that in a
7:55
little more detail. But I really wanted to just
7:58
ensure that we were building a product that could help better
8:05
understand a patient's cognitive health, help better support
8:10
patients who may be socially isolated or feeling lonely, and
8:14
ensure that we could capture their lives in such a way that
8:18
it helps preserve their legacy. OK, so if, if I hear you
8:22
correctly, what triggered you was that all of the, let's say,
8:29
information exchange was around, let's say medical things like
8:34
did they take their medication? Were they feeling tired?
8:37
Yes or no? Could they speak to you but not
8:40
around, let's say? How does this process work?
8:46
What steps do I need to take? How can I preempt maybe certain
8:50
declines that are inevitable because we can't treat them yet,
8:54
but at least we we can sort of signal them early.
8:58
Is that? Yes.
9:00
And to add, just to add to that it, I was really compelled by
9:04
this opportunity to understand what our loved ones were
9:11
encountering during the course of their day when I was absent
9:16
as their caregiver. I was absent because I'm
9:18
working. They may share some very
9:20
relevant information and relevant either contextually
9:25
from a clinical standpoint or relevant from a personal
9:29
standpoint, but I'm not there to hear it or see it or engage with
9:33
them on it. And so it so these discrete time
9:37
points that we have today, maybe it's first thing in the morning,
9:40
maybe check in on them during your lunch hour, maybe it's in
9:43
the evening after your work, you've missed a a good chunk of
9:47
their day and you really don't know what what they're
9:50
experiencing. So Cujensis gives them an
9:52
opportunity to speak whenever they want, and then you have a
9:55
transcription of what that engagement was so that you can
10:00
then reconnect with them in a more meaningful way and talk to
10:03
them about a topic that is specifically of interest.
10:07
And I wanted to build that out and help help, you know, make
10:11
that a reality. Yeah.
10:14
Well, that sounds like a really meaningful product.
10:18
Thank you. Before we go into the actual
10:23
solution that Gorgensis is building, yeah, we're getting
10:29
older because of medical advances.
10:32
So that's the good part. But does that also have a
10:34
downside like I heard you mentioned loneliness, but but
10:40
are there any others side effects of that, let's say also
10:44
maybe also comorbidities etcetera?
10:48
Absolutely. So, you know, as a normal part
10:50
of aging, we will start to see some breakdown of some of our
10:55
systems in some way, shape or form.
10:56
And it it, it, you know, we hear a lot about diabetes and
11:01
cardiovascular disease and some of those, you know,
11:04
osteoarthritis, some of these conditions that we associate as
11:10
a normal progression of aging. Perhaps the severity of each one
11:14
of those conditions obviously is attributed to a variety of
11:17
different factors, genetics, nutrition, sedentary lifestyle,
11:22
etcetera. However, what we do know is
11:26
that, you know, as we're, as we're aging and as we have this
11:31
opportunity to then, let's say, collect these comorbidities that
11:36
contributes to, you know, the, the cost associated with
11:41
treating those comorbidities and the cost of medicines, for
11:45
example, cost of physician appointments.
11:47
And then sometimes it also creates this shrinking social
11:53
connection or social circle because unfortunately, as we all
11:59
pass and in some point in time, we'll pass, you know, our, our,
12:03
our existing circle of friends can, can shrink.
12:07
And that creates this sense of, of social isolation or
12:10
loneliness. Furthermore, fewer and fewer
12:13
families are living in a joint family situation.
12:16
There was a time eonsible and not too, too long ago, where
12:19
many families lived together and that maintained that social
12:23
connectedness. Now we see fewer cultures doing
12:26
that. And you know, we, when we think
12:29
of the cognitive impact that that has, you know, the World
12:33
Health Organization estimated about $1.3 trillion was spent on
12:40
cognitive issues such as dementia and Alzheimer's in, in
12:44
2019. And that's only expected to, to
12:47
grow. And so we we anticipate that
12:50
this is, you know, the just the natural progression of aging
12:54
potential for dementia and the cost associated with that
12:57
dementia will will lead to health system burden and family
13:04
structure burden. I have two follow up questions.
13:07
One is is, is there a difference between men and women in terms
13:12
of how cognitive decline develops?
13:17
Yes. Well, so I mean it's interesting
13:18
because you know, I think in terms of how from a biological
13:23
standpoint, I think researchers are trying to understand this.
13:28
When we think about cortisol levels for example that are
13:33
higher in post menopausal women, impact of cortisol on areas of
13:39
the brain responsible for our memory, long term memory as well
13:44
as our executive functioning. So the hippocampus and the
13:49
prefrontal cortex when we see the impact of cortisol on on
13:53
these structures and therefore the impact on average daily
13:58
living activities, it it tends to create this difference in the
14:05
impact on men and the impact on women.
14:09
But preceding that is just the the statistics around how many
14:16
men and women can I speak to in in the US, how many men and
14:22
women live alone? And we see that 16.2 million
14:27
people over the age of 65 in the US live alone.
14:31
Of that 10.5 millionaire female and 5.7 are male are men.
14:40
So. It's 2 to one ratio. 2 to one
14:43
ratio of women who live alone who are then at A at a higher
14:47
risk of social isolation and higher risk of loneliness and
14:52
these people are then more susceptible to cognitive decline
14:58
you know cortisol related stress and.
15:05
Does it always also impact that, say, the timing of the start of
15:09
treatment? Because I can.
15:12
I do know from our previous conversation that you told me
15:16
something around that people are later to the game in terms of
15:24
diagnosis because they are lonely and because they're
15:27
thinking like it will. Yeah, maybe it's normal.
15:31
Maybe it's me. Yeah.
15:33
No, I thank you for raising that because it it is an issue.
15:37
I think, you know, this does affect women disproportionately
15:40
as a result of that, you know, and just in a few ways you
15:45
meant, you know, we, we talked a little bit about how people are
15:47
women are living alone. Women represent 70% of the care
15:52
hours associated with their their loved ones.
15:57
And that creates this sense of delayed health in their own or
16:05
delayed healthcare in their own health.
16:07
For example, maybe they put off going to the doctor even though
16:12
they have a symptom that they they think it's just kind of in
16:15
the back of their mind. I got to get this checked out.
16:17
I got to get this checked out. OK, I'll get to it.
16:18
I'll get to it. Because right now I'm busy
16:21
taking care of everyone else. So what ends up happening is
16:26
they, they themselves delay the appointment that in turn, you
16:33
know, potentially presents themselves in a more with a more
16:36
complex or advanced case of something.
16:40
So their diagnosis is therefore delayed and then the treatments
16:43
are that much more complex, costly and ultimately
16:49
challenging from an outcomes standpoint.
16:51
So that also answers my second question which was around the
16:56
impact of loneliness. Last question on this part is
17:01
how much do we understand of cognitive decline right now and
17:06
how important is it to get more data on that more let's say,
17:11
contextual information on how this actually works?
17:15
Yes. So cognitive decline can be
17:19
described in a in a lot of different ways and it's used
17:23
colloquially in a lot of different ways.
17:25
So for example, they have they look at age-related cognitive
17:31
changes and sometimes someone will say, I feel a decline from
17:34
that standpoint. But when we what we really want
17:38
to do is we want to focus on the cognitive decline that is
17:41
associated with changes to someone's average daily, you
17:47
know, lifestyle and our activities.
17:50
For example, you know, do they experience certain some some
17:54
memory loss or a reduced mental acuity?
17:57
These things are very important to pay attention to and are
18:01
often times just kind of pushed aside.
18:05
We pushed it aside because we're busy.
18:07
Everyone is very, very busy. And, you know, aging parents
18:13
also don't want to. You hear of them saying things
18:16
like they don't want to burden their children.
18:18
Yeah. They don't.
18:20
Want that many times? Yes.
18:22
So, you know, so then they don't mention these things.
18:25
And I think that that's a problem that we're trying to
18:27
solve for as well, because we all interpret importance and
18:30
relevance in our own way. And that that that actually
18:37
impacts how we understand and assess cognitive decline.
18:43
Because initially we need to report this information out to
18:47
our care team, our clinicians and our our caregivers.
18:51
And if we don't know that this is important, then they're not
18:56
going to know what to do. And if they're not around to see
19:00
it happen first hand, objectively, you know, they're
19:04
not going to know to report it to their clinicians and then the
19:06
clinicians are not going to be able to intervene in time.
19:09
So it's important for us, we don't really fully understand
19:14
everything that there is to understand about a decline in
19:19
cognitive health, but we are making such advancements in on
19:25
the biological side, on the medical side, we now need to
19:28
catch up on the social side and understand how social
19:31
determinants of health impact cognitive decline.
19:34
And that's that's something that we are focusing on at Cogensis.
19:37
So with that, let's now shift to the solution that you're
19:42
building. Can you describe what you're
19:45
building? Because I, I, when I heard about
19:48
it, I was like, OK, I got to speak to you.
19:51
Yeah, yeah. So, so we're actually building
19:54
an AI enabled platform that people can speak with.
20:01
So there will be a an avatar that somebody will be talking
20:06
into or talking to. And we want to engage people in
20:11
conversation just with them talking, so you know that
20:16
they're not talking only about. You know, the medication that
20:20
they are supposed to take and, and it's not just about hey, did
20:23
you take your medicine this morning Durkian or not?
20:25
Did you take your, you know, hypertension medicine?
20:28
Did you take your diabetes medicine?
20:30
Their lives are so much fuller than that.
20:33
And so engaging them, we, by engaging them, we, we hope to
20:40
alleviate loneliness, which is the subjective definition of
20:45
being isolated and also provide certain trends and patterns in
20:53
their cognitive health around memory, recall, verbal fluency,
20:59
all within this context of, of loneliness.
21:02
So that the care team and the caregivers have a better
21:09
understanding of what's actually happening to the patient.
21:13
What do they really feel? So do you then also provide
21:17
diagnostics? No, right?
21:20
We are not a diagnostic, so we are not diagnosing dementia and
21:24
we are not diagnosing Alzheimer's.
21:26
What we are doing is we are providing insights around
21:30
cognitive health that could help clinicians with their
21:36
diagnostics around Alzheimer's or cognitive decline.
21:41
OK. So it's more, let's say
21:43
qualitative data about quality of life, about how people are
21:49
feeling and with that, I guess also trend information that
21:54
you're feeding into the system, so.
21:59
What is the? Result for the, let's say the
22:02
caregivers, what can they do earlier or better as a result of
22:06
this? So let's just take a scenario
22:10
where you are busy at work all day and you call your mother in
22:16
the evening after work and you say, hi mom, how was your day?
22:21
How did everything go? What did you do?
22:25
And she responds, oh fine, you know, it's the usual, nothing
22:31
new, all's good. And you're now left to take that
22:35
at face value. Now we know our brain is, is a
22:42
wonderful organ and that it will compensate in any way, shape or
22:47
form that it can. And so if let's say, for
22:50
example, we forgot something, it may not relay that as a
22:56
significant issue. So what, you know, you forgot
22:59
about something, big deal. In the in the grand scheme of
23:02
things, is it really a bad thing?
23:06
And the answer may be not maybe it's not a big deal, but perhaps
23:12
that lack of information or the lack of recall when prompted as
23:18
a question could offer insights to something that is meaningful.
23:25
So what do I mean by that? Let's say, for example, you run
23:29
into somebody, and in the course of this conversation with your
23:32
mother, you say, how was your day?
23:34
And she says, fine. And you say, oh, great, you
23:36
know, I ran into Mr. Johansson, you know, And she just pauses.
23:45
She's like, I don't know who this person is.
23:47
And this is someone who, you know, you know, she absolutely
23:53
should know because this is someone that she has engaged
23:56
with routinely for many, many, many years, but suddenly she
24:01
doesn't recall who this person is.
24:03
If you don't know to ask the question or you don't mention
24:08
Mr. Johansson, she would not give you that response.
24:15
And so you would not have ever even noticed that there was
24:17
something missing. You would have just kind of gone
24:20
about your day and said, OK, well, great.
24:22
I'm glad everything is fine. I'm glad you're doing all right,
24:24
mom. You know, call me if you need
24:26
anything and I'll check on you in the morning.
24:28
Or the kid. My kids are fine and
24:30
everything's going well here. That's it.
24:33
And sometimes those discussions, by the way, do happen and they
24:37
they have a role as well. So maybe not to make few people
24:41
feel too guilty, but once in every while it's good to at
24:44
least drill down a bit on, especially if you met people
24:50
that you both know very well. Yes, absolutely and.
24:55
It's a good check that check. Yeah, it's a good check, right.
24:58
And then the other thing that, you know, and conversely, let's
25:01
say, for example, she was thinking about, you know, maybe
25:08
a childhood experience, a trip that she had taken with her
25:11
family. And for whatever reason that
25:13
came up in, in her thoughts that day, but you would not have
25:20
known that because she may not think that's important enough to
25:23
share with you. So she may, using cogensis, have
25:29
a conversation, you know, because cogensis may prompt her.
25:32
Can you recall a time in your life that you had gone on a trip
25:34
or something of that nature that was very memorable.
25:38
And she might answer that question in that moment and say,
25:40
Oh, yes, I remember we went to Bruges, for example, a lovely
25:44
time. OK, well, now when you come home
25:48
and you see the dashboard that you would have access to if she
25:54
gave you access to this dashboard, you have access to
25:57
this. And you see in the transcript
25:59
that she was recalling a trip she took with her family to
26:03
Bruges when she was 22 years old.
26:06
And you may be able to engage her at that moment and say, mom,
26:10
tell me about this trip that you took.
26:14
What was that like? Where did you go?
26:16
What did you see? Did you go to the Abbey?
26:18
Did you go? Where did you go to the center?
26:19
What did you know what? Tell me more about it.
26:23
And that might spark a very different conversation with your
26:26
mom beyond the oh, I'm fine, Nothing new today.
26:30
Everything was, you know. And also sort of makes people
26:36
feel hurt. They they're hurt and.
26:39
And and and and valued. Yes.
26:41
And alleviating that loneliness, right?
26:43
So they understand that you're connecting with them, you are
26:46
making a more meaningful connection with them as a person
26:49
and not just as your mother that you're taking care of.
26:52
Yeah. So this, this solution, it's AI
26:56
based. How does it differ from let's
27:00
say sort of a chat bolt or well, reminder app, That's sort of
27:06
clear, but is it a sort of chat function or how does this work?
27:11
How does the interaction with the lonely person?
27:16
Yeah. So it's not necessarily just a
27:19
chat bot, right, Because there are a lot of analytics that go
27:22
on in the background. We do prompt you to engage with
27:26
the with with cogensis, and obviously the more you use
27:31
cogensis, the more comfortable you become with the
27:36
conversation. It doesn't feel contrived.
27:38
Initially it might feel very weird, but.
27:42
Yeah, it's the same with Hey Siri on an Apple.
27:45
The first time you use it, everyone was probably like, OK,
27:48
I wonder what what is going to happen now?
27:51
And now everyone is doing. It and it's, it feels natural
27:54
almost, you know, so you know, we have intentionally used an
27:58
avatar instead of a human form because we, we don't want there
28:03
to be unhealthy connections with this.
28:06
We don't want them relying on this system as if it was an
28:11
actual person. They need to be aware that this
28:14
is not a physician that they are talking to as an example, but
28:19
that this is part of their care team now that they can have a
28:24
conversation with journal talk about anything that they want to
28:28
talk about. However, on the back end, unlike
28:32
a chat bot who may be just kind of recording an answer and then
28:35
just spitting out an answer. We I'm sorry asking prompted
28:39
questions. We are looking for those trends
28:42
on the back end to see whether or not there's a change in
28:46
verbal fluency. How much, how many words are
28:49
they actually using to respond to a question?
28:54
How long are they thinking about the response before they give
29:00
you a response? Is that considered consistent
29:03
with what we would expect for someone of their age and
29:06
condition, or is it an outlier in some way, shape or form?
29:10
So these types of it it can detect that people all of a
29:16
sudden start taking longer for their answers.
29:20
Which is. Sort of delta compared to and,
29:27
yeah. In that population or in that
29:29
subpopulation, like what would we expect in terms of a normal
29:34
pause before we respond? Do we see any frustration?
29:39
You know, is there, is there any sense of frustration?
29:41
Because that's very telling as well.
29:43
Is there sort of sensitivity or how do you say sentiment and
29:47
analysis building? There, there will be, yes, a
29:51
little bit of, you know, sentimental analysis and
29:53
understanding, You know, what is this person's emotion?
29:57
How are they responding? You know, semantically,
30:00
syntactically, from an auditory standpoint, we're looking at
30:03
listening to see, to hear whether there is frustration in
30:08
the voice or not. If they express concerns around
30:14
being lonely, you know, if they don't say this because they
30:17
don't want to burden their children, but they feel safe
30:20
talking into a computer who they don't necessarily think is, you
30:25
know, being burdened per SE, You know, the, the, the quality of
30:29
conversations may change. So we think it's, it's much more
30:33
than what the chat bot provides in terms of functionality.
30:38
And it's much more than a medication administration
30:41
reminder. Because remember, people are so
30:45
much more than just the medications that they're on.
30:47
That information is important. And we we don't.
30:50
And it's not to say that that won't be part of the solution
30:53
that is there, but it's so much more than just that.
30:55
OK, I was wondering about one thing.
31:00
Because yes, it's incredibly intelligent to have that
31:07
conversation even and, and to, to sort of being able to listen
31:14
to people what they have to say and maybe engage in a
31:17
conversation even if there's no other human in the room.
31:22
Can this, the word that came to my mind is how do we avoid that?
31:28
This is seen as spying or snooping on people.
31:32
Right. And you know, this is a, a
31:36
challenge, I think for all passive listening devices,
31:42
right? And sort of virtual assistants,
31:47
etcetera. Over time, you know, I think
31:52
over time you start to build trust with Siri or with any
31:59
other virtual assistant and, or application.
32:03
And the, the, it's, it we need to be responsible in terms of
32:09
what, what is it that we're asking?
32:11
What, why are we asking certain questions?
32:13
What information is going to be relevant?
32:17
And perhaps they will find that talking into a computer system
32:22
is strange. They may initially think it's a
32:25
little intrusive. However, realistically, if we
32:31
take an example of someone living in a care facility as an
32:36
example or a care community, they are often asked questions
32:42
by staff, by their children, by others, and they may feel that
32:50
those questions are a little intrusive.
32:54
For example, my mom, who again, she, she, she, she, she became a
33:00
widow somewhat unexpectedly. And when we started asking her
33:09
questions, you know, how are you doing?
33:11
How are you feeling? What are you eating these days?
33:14
How, how are you sleeping? Did you shower?
33:17
You know, we, we were trying to, we didn't want to come out and
33:20
say, were you depressed? You know, because we knew that
33:24
would trigger her in other ways, right?
33:25
Like, no, I'm not depressed. My mother is a very strong
33:28
woman. And you know, she does not want
33:32
to feel helpless or feel dependent on anybody.
33:39
She she values her independence. And so initially when we were
33:43
asking her these questions, she sat us down and she says, I do
33:46
not want to be coddled. I do not want you to ask me
33:49
these questions. It is, you know, I'm eating
33:51
fine. It's none of your business what
33:53
I eat, what I don't. I mean, she doesn't drink
33:55
alcohol. So we were not worried about
33:56
that, but we were just more worried about her cocooning
34:02
herself and just not engaging with anybody and not eating and,
34:06
and not taking her medicines and all of that.
34:09
And she just kind of said, you've never asked me these
34:11
questions before. You never asked me if I took my
34:13
medicine or not. You assumed I took my medicine.
34:15
Why would I not take my medicine?
34:17
Now? Stop doing that.
34:18
She was very angry initially, you know, and so she felt that
34:22
that this human engagement was intrusive.
34:25
But for whatever reason, you know, when she talks to, she
34:31
actually talks to Siri and it's like, I don't know, I, I feel
34:33
like she, she, she just kind of gets it off her chest perhaps,
34:41
but doesn't feel like she's burdening anybody while while
34:44
doing that. And so it doesn't feel as
34:46
intrusive. Yeah.
34:47
Is it then also something around that it's sometimes easier to
34:53
talk to about emotions to a complete stranger?
34:57
Stranger totally. I was just about to use the
34:59
example of those people who sit next to you on the airplane and
35:03
tell you their entire life story in the 1st 15 minutes of a 10
35:06
hour flight and and you know, the longest time for the longest
35:12
time was like, why would anybody do that?
35:13
And then it, and then I realized that you're, you're sitting next
35:17
to somebody for 10 hours and you just want to know who that
35:20
person is or they want you to know who that person is.
35:22
Like I'm not one to share my life story with the person next
35:25
to me, but I'm often the person who is on the receiving end of
35:28
someone's life story. And then it goes into and by the
35:32
end of it, I feel like sometimes I'm a a marriage therapist, but
35:36
you know, the conversation, but, but, but I think it's all about
35:46
this, this need to feel socially connected, right?
35:50
And so I really think that that's at the end of the day,
35:53
what it, it is, the is, what is the root?
35:57
What's it all stems from? So trust is a big factor for for
36:02
elderly people, especially for elderly people.
36:07
And also what you see is that they're, let's say, less
36:11
inclined to adopt tech. How or at least be more
36:16
skeptical around it. How do you deal with that?
36:20
Yeah. Well, as as they should be,
36:22
right, as we all should be because it's a, you know, a
36:24
trust but verify kind of a situation we have to win their
36:28
trust. And you know, we intend to do
36:33
that by keeping the, the usability of this product very
36:39
simple. The the trust and the technical
36:45
literacy is very important. Now, from an aging standpoint,
36:51
each tech seems to fall into a single category as someone,
36:58
anyone over the age of ironically 55 or over, which to
37:02
me does not feel like very old. You know, I personally don't
37:07
like to be put in from a, from a tech savviness standpoint, don't
37:11
like to be put into the same category as let's say my 92 year
37:14
old father-in-law, right, because I'm involved in tech,
37:17
I'm working with tech. And so you'll have, you know,
37:20
senior, you'll have folks who are over the age of 60, let's
37:23
say, or 60, you know, 62 maybe even who are executives within
37:28
large organizations and large corporations.
37:31
And they are engaging with tech all the time.
37:34
Now, are they in the back end writing code for AI platforms
37:39
and machine learning algorithms and RPA remote, you know,
37:42
robotic processing algorithms and so forth?
37:44
Maybe some of them are, but not everybody is doing that.
37:47
And so, you know, when we think about AI agents and so forth,
37:50
maybe they're not engaging with AI agents to the extent that
37:53
everybody is, but they may have more knowledge of it than let's
37:58
say my 92 year old father-in-law does.
38:01
And he's adorable. He's a a wonderful, wonderful
38:04
person and he's curious. So we have to win the trust of
38:11
people who are within this spectrum.
38:15
People who like my father-in-law, may be naive
38:18
about technology and may be interested in learning more and
38:24
be less critical. And they may, he may actually
38:28
trust the technology more inherently because of that and
38:32
others who are, let's say, working professionals who are
38:35
now retiring, who have knowledge of technology and are perhaps
38:39
even more skeptical because they know of the potential negative
38:44
ramifications of, of technology. So it's, it's up to us to to win
38:48
their trust. And I think we're going to do
38:50
that effectively by focusing on the usability of the product and
38:55
keeping it simple to use and also by having very specific use
39:00
cases around, you know what, what problem we're trying to
39:04
solve for and, and namely cognitive health insights and,
39:09
and loneliness. OK.
39:11
So the question that I forgot and now that came back is we
39:16
talked for quite a while and just now about your personal
39:20
experience, especially with your parents.
39:24
How did that does that help you in the development of the
39:28
software? It helps me tremendously, helps
39:33
me tremendously because I see myself as a user of this.
39:38
So I can take not just hypothetical scenarios, but I I
39:46
really understand the, the journey that the user will take.
39:51
So what is the workflow, if you will, the patient's journey, the
39:56
caregiver's journey, all of these different personas that we
40:00
will then be using in order to as as users and segmenting as
40:04
users and really understanding what that feels like, what it
40:12
felt like, what it's like for the surviving parent.
40:18
And, and you know, for, for the caregivers and then what
40:21
information do I want to relay to the care team, her care team
40:26
or his care team? It gives me that first hand
40:30
experience. And so I'm really, I consider
40:35
myself very fortunate to be part of this team and to be solving
40:38
this this problem. Yeah.
40:41
And I think there's a long way still ahead.
40:45
So I want to switch a bit to from the current state and the
40:52
problem you're solving in the near future to let's say later
40:57
stage. In our prep call, we talked
41:00
about first you, you, let's say, deal with loneliness in elderly
41:07
people, people in nursing homes. But we also discussed the impact
41:12
loneliness may have on veterans, on athletes, on maybe other
41:17
groups. How do you see cogensis evolving
41:21
from here? I think that cogensis,
41:24
especially as we start to understand some of the cognitive
41:27
health insights is really well positioned, will be really, I
41:30
should say, will be well positioned to help these other,
41:35
you know, groups who may be experiencing either loneliness
41:41
and or dementia like symptoms and signs.
41:45
You know, CTE is, is a, is a big issue in, in athletes who have
41:52
repeated injury to the point where they developed this
42:00
encephalopathy that resembles many times, presents itself many
42:04
times as a cognitive decline. And, you know, just a lack of
42:09
executive functioning, changes in behavior and mood and so
42:12
forth. And so we can easily see an
42:15
opportunity to engage with with professional athletes who may be
42:21
at risk for CTE as an example. So maybe maybe could you
42:27
describe what CTE is for the non medical medical schools among
42:33
us? Yes, absolutely.
42:35
So CTE is considered chronic traumatic encephalopathy and
42:41
this is a result of repeated impact to the head.
42:47
And we can see this in, you know, potentially like rugby
42:49
players or hockey players or football players.
42:53
And you you hear about it more in in the NFL, for example, but
42:57
there are a wide variety of sports that where where people
43:03
incur these injuries and. Yeah.
43:05
Also soccer. Yeah.
43:07
And soccer. As well, people are soccer
43:10
players are heading the ball and often with quite a bit of impact
43:16
it has on the on the skull, but also on the brain.
43:19
Yes, exactly. And so the, this, this condition
43:23
leads to many times dementia. And so once we start to see the
43:28
parallels of cognitive decline or cognitive health and provide,
43:33
you know, we have an opportunity to provide insights into that
43:36
group of, for that group of people as well.
43:39
And hopefully change behavior and maybe even Dr. innovation in
43:45
in gear and tools that could be used to in protocols that could
43:49
be used to, you know, and rule sets.
43:54
Yeah, absolutely. Policies, procedures, protocols
43:57
of how to treat patients or you know, who may have and
44:03
concussion protocols, for example, when I said protocol,
44:05
concussion protocols, things of that nature and what to look
44:09
for. It's a lot of this is patient
44:11
education and family education too.
44:12
What do you look for? So somebody is, you know, let's
44:15
just say someone is agitated and you know, they, it's like, why
44:19
are you agitated? OK, Is it, is it 'cause you're
44:21
bored because you're no longer playing a sport that you love?
44:25
Or is there something else that's going on?
44:27
Yeah, who knows, OK. For, for how would this work for
44:32
veterans? Is that then more in the PTSS
44:36
scope? So, so, you know, for veterans,
44:39
those who are repeating or recalling a, a, a negative
44:44
memory of, you know, something that happened in, in, in
44:48
wartime, they, they may feel that they want to socially
44:54
isolate. Either they're embarrassed or,
44:57
you know, they we, one of our Co founders is a, is a veteran to
45:02
tour of duty in Afghanistan and our COO, you know, and, and then
45:08
speaking with other folks who are veterans, these are people
45:12
who have seen a lot have endured a lot may suffer from some form
45:19
of, of anxiety, may from some form of depression, may in turn
45:25
isolate themselves as a result. And when they isolate
45:28
themselves, they're at a much greater risk for developing
45:32
dementia. You know, loneliness alone has
45:35
shown, like according to the American Academy of Sciences 317
45:39
page report, loneliness, you know, will increase has a 59%
45:43
increase risk of functional decline.
45:44
Let that sink in. That's huge.
45:47
Well, that's huge. And so it's not just about
45:50
people who are already diagnosed with dementia or already
45:53
diagnosed with Alzheimer's there.
45:55
There is a social component to this that is putting people at
45:59
risk for all of these other medical conditions and outcomes.
46:03
The odds ratios are incredible. How can cogensis deploy AI to
46:09
serve humanity in the long run beyond the the the cognitive
46:14
capabilities and signaling diseases like Alzheimer, like
46:19
dementia? Yeah.
46:20
So I think AI in general has great potential if it's
46:25
developed and applied correctly. I think that applications like
46:33
cogensis that are geared to improve loneliness and or social
46:41
connection have an opportunity to provide those insights around
46:46
a myriad of conditions, not just cognitive decline.
46:52
And so there is an opportunity to improve care management for
46:56
patients, to help clinicians with data and insights that will
47:06
hopefully reduce their burden and burnout.
47:10
We've all heard of the extreme burnout that our clinicians who
47:15
are just such integral, integral members of our society are
47:20
encountering and facing as a result of the demands placed on
47:24
them. So there is an opportunity to
47:27
help them provide them those insights so that they are not in
47:34
turn so taxed that they are making clinical errors in or
47:41
clinical judgement errors that has an impact on overall
47:47
outcomes for for people of all different diagnosis and
47:53
conditions. So when we're thinking about
47:54
humanity, humanity is such a big scale, right, That, you know, if
48:00
we want to solve World Peace and solve a humanistic problem, you
48:04
know, we, we, it's going to take time to, to, to get there.
48:09
Cogensus, I think alongside other AI platforms, has great
48:15
potential if it's developed correctly.
48:17
It has the right governance, pays attention to ethics,
48:20
ensures that you know benefits outweigh risks.
48:24
How do you do that? How do we do that?
48:28
So we have to have a very clear understanding of what our use
48:32
cases are, know what our swim lanes are, know what problem
48:36
we're solving for, avoid overreaching in the sense that
48:44
you know, we don't we we want to take calculated risk, but we
48:47
need to understand how the risk compounds.
48:51
And so that it's not just transaction for transaction, but
48:54
how does each risk within each transaction compound to a point
48:59
where it may out weigh the potential benefit to the
49:03
patient. So for example, if the benefit,
49:09
the potential benefit to the patient is that we are able to
49:12
provide insights to clinicians, the potential risk is that the
49:19
information that is provided is either not applied in the manner
49:28
in which it was intended or is unused all together.
49:35
So you know, So what what's the impact of that?
49:39
And so if it is not applied as intended, then we have to
49:45
educate the and have to design so that the, you know, the
49:49
clinicians understand from a human factor standpoint, they
49:52
know what to do with the information.
49:55
Same thing goes for caregivers, same thing goes for the patients
49:57
themselves. We want it, we want patients to
49:59
have agency over their own life. And when we say we want people
50:02
to age with dignity and age in place with dignity, we want them
50:05
to have that agency. And so we hope that this will
50:08
provide that benefit is agency. Yeah, and I think agency is
50:14
something we all want to have until we die.
50:18
Yes, absolutely. So we have to think about this
50:21
responsibly. What are we doing so that we can
50:23
provide the right information at the right time so that
50:27
clinicians and and caregivers can intervene appropriately when
50:31
appropriate and can leave them alone, leave people alone when
50:35
they don't want to be, you know, interrogated constantly.
50:39
So it's the balance. It's finding that balance.
50:41
But, you know, overall, I think that, you know, we just need, we
50:48
need to make responsible choices.
50:50
It's an evolving field. We're not there yet.
50:53
This is very nascent in, in its, in its applications.
50:56
And you know, we, we need to just make responsible choices as
51:00
a human community and not just simply driven by individual
51:04
ambitions. So.
51:07
So how, how do you balance that? I mean, how do you avoid that?
51:13
It's becomes driven by personal ambition rather than solving the
51:18
the bigger issue is there some sort of is that done in, in
51:23
terms of governance, in terms of of how information is protected?
51:28
Right. Well, I mean, I think it's a
51:29
combination of all of them, right.
51:31
If we look at the World Health Organization and some of the
51:33
agreements that they have in place to help solve global
51:37
health issues, this is going to take, this is a, if you're, if
51:43
we're still speaking about humanity as a whole, it's a huge
51:48
undertaking and it cannot be solved by one.
51:51
It should not be solved by one person alone.
51:54
It, it needs to be solved by, you know, by, by, by all members
52:00
of, of, of the community, right and representing all members of
52:03
the community. So understanding voices that
52:06
need to be heard, having an empathetic approach to
52:11
understanding, but having an empathetic approach to
52:14
approaching cognitive decline or some patient's condition,
52:19
understanding what benefit looks like and agreeing to that,
52:24
having, having, you know, a charter in place that agrees to
52:30
the better good, the good of the the community over 1's
52:35
individual ambitions, I think will be important.
52:37
It's hard to it's hard, it's hard to really fight against
52:40
human nature, right? Because humans are are are
52:45
selfish by nature, I think. And so it's hard to kind of
52:47
fight that. You have to really want to solve
52:50
a problem for a community. And that's what we're trying to
52:53
do as as as broader ecosystem. Yeah, so I want to talk about
53:00
the, the, the big thing of humankind and and selfishness is
53:07
money. For a bit.
53:10
You are going to be part of an accelerator soon.
53:15
That's the start of of your fundraising journey as well.
53:22
So my podcast started when I read about two percent, 2% of
53:27
funding that goes to female founders.
53:31
I was wondering what your journey has been like as a Co
53:35
founder of a Med tech startup and getting funding in getting
53:41
accepted into this accelerator. Yeah.
53:44
We are so excited to be part of the Caduceus Health Accelerator,
53:49
Caduceus Ventures Accelerator. From what we are told, this was
53:54
a very competitive process and you know, and so we're delighted
53:59
to be part of this cohort. And so in terms of fundraising,
54:06
in our experience, we are in an economic time that is a bit
54:13
challenging. Equity is expensive right now.
54:15
People are are are a little bearish at the moment, but we
54:19
hope that this will be changing soon.
54:23
What we, if we had, you know, what we've been told over and
54:27
over again is that, you know, the VCs that we've spoken with,
54:35
angels that we've spoken with, family offices that we've spoken
54:39
with, love the concept, they love the product, they love the
54:43
team. And now it's a a matter of
54:46
timing. We want to, we're raising $1
54:49
million, which is not exorbitant so that we can build out this
54:55
platform. We have the team in place,
54:57
including the development team. We have letters of interest from
55:01
dozens of of entities including state governments, providers,
55:08
health systems, payers, senior communities, etcetera.
55:14
We would like to continue engaging with these investors to
55:20
invest in age tech and see the opportunity that we that that we
55:28
present the problem that we're solving for the solution that
55:31
we've created and and the benefit to our customers.
55:35
OK. What kind of investors would you
55:38
like to attract? Because Med tech is difficult
55:42
anyway, right? You're in software, so maybe
55:45
that's a bit easier and it's AI well.
55:49
We're, I mean we are honestly, we're, we are working with
55:53
health specific investors, we are working with generalists as
55:56
well who have more of an impact who want, you know, impact
56:00
investing and are interested in social impact.
56:03
So I would say health people who are interested in solving a
56:06
healthcare problem and people who are interested in making a
56:09
social impact, those are the two that we tend to attract and, and
56:17
so those are the ones that we've been speaking with.
56:19
But this has broad, you know, broad reach.
56:22
There's a broad reach here and broad scale implications of, of
56:25
building out an AI platform like cogensis.
56:28
So the opportunity exists for a variety of different investor
56:34
types, if you will. So and and have you been
56:38
involved in in the pitching process?
56:41
Yes. So what is your experience
56:42
there? The experience has been very
56:45
generally very favorable. What we have found is because
56:48
we're pre product, most of the conversations have been kind of
56:54
like this. We love the idea, we love the
56:56
team, we love the product. You're a little too early for us
56:59
because we tend to invest in Series A or Series B or maybe
57:05
even seed stage funding. So we think you need to be
57:09
involved with an accelerator or you need to go through, you
57:12
know, some precede Angel funding, etc.
57:16
And then once you have that, then go ahead and reach out to
57:19
us again. And so we actually have a, a, a
57:22
short list of VCs who have asked us to contact them.
57:26
We also have a short list of technology companies that have
57:30
asked us to contact them again because they are interested in
57:33
integrating our product into their technology platform.
57:37
And so, you know, they have actually been reaching out to us
57:41
just to kind of keep tabs on, on the progress that we're making.
57:44
So we're hopeful, very hopeful that this admission into the
57:47
Caduceus Venture Accelerator is going to catapult our our
57:54
progress and and enable those conversations to to move
57:58
forward. Well, I hope so too.
58:01
Any tips for other female founders who want to try?
58:07
Yes, do it. Just do it.
58:10
Get out there and, and, and build something.
58:15
There is a need to have the female perspective in the work
58:20
that is being created. I can speak from a healthcare
58:24
standpoint because that's my career, but I can see this, you
58:27
know, female founders in every single industry have an
58:31
opportunity to go out there and do it.
58:33
Don't be afraid. Sounds like a great place to
58:36
move to my last question, which is I'm on a mission to make
58:41
inclusion in tech the new normal by the end of 2032.
58:46
Based on your experience, what is the one bold move we all need
58:51
to make to get there? Yes thank you for this question.
58:56
Honestly, I feel like we, we just need to, and I use the word
58:59
just in air quotes, 'cause I realize this is not a just, but
59:04
shift our mindset to allow inclusion.
59:08
I think that the framework or the process of innovation isn't
59:13
a sex based trait or a culturally driven issue.
59:18
We identify a problem we create. We want to create a solution
59:23
using calculated risk to solve that problem and we want to
59:26
benefit our customers. That's the process.
59:30
It doesn't matter whether that is a, it shouldn't matter,
59:35
right? From a process standpoint, it,
59:37
it's the same. So, so that's right.
59:43
And so, you know, the, the, in my opinion, the issue is really
59:47
just changing our perception and getting out of our own way.
59:53
We talk about failing fast and innovation.
59:56
And you know, if this is an accepted principle, then why
1:00:01
should we place caveats on it based around?
1:00:04
The sex or the ethnicity of the team, what difference does it
1:00:09
make right if if you are going to apply a basic principle of
1:00:14
failing fast in innovation because you have an innovation
1:00:18
that it because you've identified a problem, you have
1:00:22
created a solution using calculated risk and you are
1:00:27
potentially benefiting a customer and will produce good
1:00:30
outcomes. What difference does it make if
1:00:34
the founder is female or male? If that's the only factor, then
1:00:39
I think perception is the issue. OK, so change to perception and
1:00:43
also change then change then our own mindset.
1:00:48
Yes, it's a change in my opinion.
1:00:50
We need to shift our mindset to allow for inclusion, full stop.
1:00:56
Sounds like a great place to close Kathy Joshi.
1:01:00
Thank you so much. Yeah.
1:01:01
Thank you for being a guest and thank you for the wisdom and for
1:01:05
sharing what the wonderful work that you do with Cogensis with.
1:01:11
US. If people want to learn more
1:01:14
about you, or connect with you, or learn more about cogensis,
1:01:19
where would you like to send them?
1:01:21
Well, you can learn more about cogensis at at cogensis.com,
1:01:25
www.codensis.com. You can reach me by e-mail at
1:01:33
raju@cogensis.com and I would be delighted to hear from your
1:01:37
listeners. I'm very grateful to you for
1:01:43
what you do and and for driving this mission that you have to
1:01:52
make inclusion in tech the new norm.
1:01:54
I think you said by the end of 322032.
1:01:57
Thank you for doing that. I know you're busy and I know
1:02:00
you could have done 100,000 other things, but I I appreciate
1:02:04
and I'm grateful that you've felt this was an important
1:02:08
enough issue to take it on. So thank you for doing that.
1:02:13
You're more than welcome. We'll include all the links in
1:02:17
show notes. So once again, Raju Joshi, thank
1:02:20
you very much for being my guest.
1:02:22
Listeners, thank you very much for listening.
1:02:24
And until the next episode, keep being awesome.
1:02:31
Well, that's it for episode 112 of Women Disrupting Tech with
1:02:36
Raju Joshi on how Gogensis uses AI to tackle loneliness and
1:02:41
cognitive decline. Rajo and I hope that you found
1:02:45
our conversation as inspiring and insightful as we intended it
1:02:49
to be. And if you have any questions
1:02:51
for Rajo, please leave them in the comments on Spotify, Apple,
1:02:55
or YouTube. I'll make sure you get an
1:02:58
answer. If you'd like to learn more
1:03:00
about Rajo and Cogensis, there are links in the show notes.
1:03:05
And as usual, you can find the show notes at Women Disrupting
1:03:09
Tech dot blog and that link is also in the description.
1:03:13
If you've enjoyed this episode, please consider leaving a rating
1:03:17
or a review on Apple or Spotify. Leaving a rating or a review
1:03:22
helps new listeners find podcast and supports my guests by giving
1:03:27
them a bigger platform. If you're a regular listener,
1:03:31
please make sure you follow Women Disrupting Tech on
1:03:33
Spotify, Apple, Amazon or YouTube.
1:03:37
Following Women Disrupting Tech is a simple 0 cost way to
1:03:41
support Raju and my older guests in making inclusion in tech the
1:03:45
new normal by the end of 2032. If you'd like to learn more
1:03:49
about Women Disrupting Tech you can follow us on social media.
1:03:53
You can find us everywhere at Women Disrupting Tech.
1:03:57
There's also Subtech page for Women Disrupting Tech.
1:04:00
On there you'll find in depth stories and behind the scenes
1:04:03
insights from each episodes, the trends that I've spotted across
1:04:08
120 plus conversations so far. So don't forget to subscribe to
1:04:14
women disrupting tech on sub stack.
1:04:17
Finally, if you want to connect with me, please reach out via
1:04:21
e-mail. Dear Hyun at women disrupting
1:04:24
tech dot blog or connect on LinkedIn.
1:04:26
Just search for my name and you'll find me.
1:04:30
With that, I wish you a wonderful day.
1:04:32
I hope to see you in the next episode and until then it being
1:04:36
awesome.