Setting The Stage: AI Meets Therapy

Dr Andrew Greenland 0:00

Welcome back to Voices in Health and Wellness. This is the podcast where we explore what's really working in today's rapidly evolving health landscape from functional medicine and mental health to tech-driven solutions, reshaping the care experience. Today's guest is someone doing truly transformative work at the intersection of psychology, AI, and access to care. I'm joined by Dr. Ayelet Hirshfeld, a clinical psychologist, psychoanalyst, and the founder of DiversitUS Psychotherapy & Psychoanalysis Inc. She's also the driving force behind PsyEcology Inc., an ethical AI platform tackling one of the biggest challenges in therapy, client dropout. So with that, I'd like to welcome you to the show. Thank you so much for joining us today.

Dr Ayelet Hirshfeld 0:44

Thank you so much. It's a pleasure to be here. So if as you've mentioned, I am the founder and CEO of PsyEcology Inc. It used to be called your AI Psychoanalyst, which just changed the name to be more inclusive. And for a decade I've worked clinically, I work in advisory capacities and in the intersection of psychology, psychoanalysis, trauma, technology, and human behavior. And I am very, very um interested in the clinical reality. So what brought me here is not really an academic background, but more sitting with patients and realizing again and again that therapy doesn't usually fail because people don't want help. But it often fails because the system doesn't support getting the right person to the right clinician at the right time.

Dr Andrew Greenland 1:46

Got it. Okay. If we could perhaps rewind a little bit and just maybe start with your sort of journey to where you've ended up and what sort of inspired you to do this work and specialize in the psychoanalysis and the psycho specialities that you do, just so we have some context for who you are, where you've come from.

Dr Ayelet Hirshfeld 2:05

Yes. So I've

Ayelet’s Path And Systems View

Dr Ayelet Hirshfeld 2:07

earned a PhD from Alliant International University. I've completed a postdoc at Stanford and my analytical formation at the Lacan School of Psychoanalysis, as well as the healthcare executive program at UCL's Business School for Health. And that really allowed me to zoom out and be able to look at the system as a whole. Really, with a decade of experience providing care to patients, working in different systems, both in private practice and community mental health, and I was really able to look at the difficulties, challenges that mental healthcare is facing, and be able to integrate and think about how AI can help transform and better patients' experience and clinicians' experience at the same time.

Dr Andrew Greenland 3:06

Thank you. Really helpful for that context.

The Dropout Crisis Explained

Dr Andrew Greenland 3:08

So let's start with the challenge. You just alluded to earlier on. Why is early dropout such a persistent issue in therapy? I think you mentioned about the matching and it's you know getting the right therapist. Are there any other aspects to this?

Dr Ayelet Hirshfeld 3:21

So in practice, um event usually it shows up as you know, people starting care with motivation, and then they struggle to continue. Not because therapy can't help, but because the fit isn't right, the process itself is too fragmented. And so if you think about it in the medical field, imagine you broke your hand and you're then matched with a gynecologist. And so there are doctors that and you really need help, but these doctors do not specialize in the care that you need. And so many people cycle through multiple therapies before something clicks. And each time they have to retell their story, open difficult material, rebuild trust. And for a lot of people, especially during crisis, that's not sustainable. And then on the other end, there are the clinicians that are overwhelmed with screening, intake, administrative work, and coordination that takes up a lot of time and energy and contribute to burnout and workforce train. And so this problem lies on both the client side and the clinician side, and this is what really led me to fund Psychology Inc. And as I mentioned, you know, before both private practice and large systems or community settings, in private practice, dropout is generally lower than what we see in large systems or community settings. And that's because clinicians have more autonomy, they can assess fit informally, they aren't thrashing intake, and continuity tends to be much better. And the psychologist is responsible for scheduling the sessions, clients that are seeking care in private practice are more knowledgeable about their needs and the particular qualifications of the therapies they're seeking or the conditions they're seeking treatment for. Dropout becomes a major crisis at the system level, especially in public system, insurance-based care, large platforms where matching is seen, intake is rushed, and people are treated interchangeably. And so then scheduling is really handled by dedicated team who are not familiar with each clinician speciality or the client-specific needs. And this is a major problem. And I can go more into the numbers both in the US and the UK in terms of the burden and the really significant and estimated burden each year annually on the government spending.

Fit, Attunement, And ANA

Dr Andrew Greenland 6:02

Yeah, I'm really interested in that burden. And um, I just wonder how much of this is a personality of the client versus the practitioner, and how much is it that is the wrong approach, or is it a combination of the two? Or sometimes the personality find it is just the approach that's wrong in your experience when you were searching this to come up with the platform.

Dr Ayelet Hirshfeld 6:22

So we know that um in psychology as a whole, in psychotherapy, there are many variables that we cannot really, you know, where we cannot really fully explain or predict who will do better and with which therapist. There is the um on the end of the qualification, you know, there there are specific issues, problems the clients present with, and where clinicians basically do not are not qualified to treat. The other end, there is, as you said, the personality of the clinician, the personality of the therapist. And we do know that 50% of the variable that we cannot account for is exactly that attunement and that personal connection to a therapist that you can relate to, which is why in psychology um we created an AI-native mental health platform that streamlines therapist client matching, and which we call ANA, and it's composed of two core components. One is a therapist client matching algorithm, and the other is a virtual therapist persona agency system. And so basically, that addresses that particular issue because it builds therapist personas that mirror each clinician's style and enable clients to interact with the therapist persona agents before booking. And it is a built to enhance the human expertise. So, again, unlike static platforms like Psychology Today, Zoc Doc, we offer a dynamic AI-driven engagement, and which is not a listing only of clinicians and expertise without the client therapist fit optimization. And so this personalized AI matching therapist persona agent allow this continuous engagement and they really do focus on reducing the friction at entry into care instead of relying on explicit self-report, such as what kind of therapies do you want, it really incorporates implicit signals of how people communicate and respond to structural feedback, pace, tone, and relational style. So this is it will give you a sense of how that clinician relates, how that clinician communicates. And because these therapist personas are trained by each clinician and they reflect the clinician's real style and areas of expertise. Fascinating. And so it's it it's the purpose is not to simulate therapy, but to let clients get a grounded sense of what it would feel like to work with a therapist before booking.

The Staggering Burden In US And UK

Dr Andrew Greenland 9:12

Do you have any sense of um we we're talking about the bird, come back to the burden again? How how many sessions is it taking clients to realize in the old sort of system that they're with the wrong therapist? Does it take more than a couple of sessions? I mean, do patients, clients spend you know months in therapy before they realize this is not a good match? Because I can see this being a huge problem that you know your thing could solve. But I just want to really get some sense of what that burden really looks like.

Dr Ayelet Hirshfeld 9:40

Yes, definitely. So this is this goes really into the heart of the problem we're trying to solve because this unmet need is really enormous, both in the US and the UK. So if you look at the United States, around 59 to 61 million adults meet criteria for some form of mental illness in any given year, but only about 52% receive treatment. So it basically means that half of the people who need care are not getting it. And even among those who enter treatment, dropout, as you say, especially early dropout, remains really common. And usually it would happen within one to two sessions that 50 to 65 percent of clients will drop out of therapy. This is a staggering number. And again, we're talking about large systems, large health systems. Um and it's a systematic problem more than a private practice problem. And this leads to, for example, out of the total US health spending of 4.9 trillion, the total mental health spending is 225 billion annually. And the indirect economic burden has to do with you know, loss of productivity, disability, unemployment, and medical comorbidity is estimated at 300 to 500 billion. So this is the problem that we are trying to solve, and which is indirect economic burden tied to individuals not receiving mental health treatment when they need it. And the numbers in the UK are one in four people experience mental health problems in any given year. And over two million people are waiting on waiting lists for NHS mental health services. Many wait for months before they start receiving treatment. And again, this leads to significant indirect economic burden of about 100 to um 150 billion pounds annually that has to do with productivity losses and quality of life impacts.

Dr Andrew Greenland 11:59

Wow, it's a very worrying statistic.

Building Therapist Personas

Dr Andrew Greenland 12:02

So talk me through the platform from the perspective of a new practitioner who's registering on the platform. What does that whole training thing look like? How does the platform get to know what they're like, how they operate, and create a persona?

Dr Ayelet Hirshfeld 12:15

So the clinician will interact with a client simulator to create their virtual therapist persona. And that will be based on, you know, there is proprietary information there that I cannot go into the details, but they will train a virtual therapist persona that showcases their knowledge, their expertise, their relational style, and it will allow clients to interact with. And clients that will join the platform will be matched based on the therapist client matching algorithm, and will basically be introduced to therapist personas that meet the criteria of the problem they're seeking help with, and will be able to choose in a very precise way a therapist that they can relate to, a therapist that really does meet the criteria for the problems they're presenting with, that has the qualifications. And um again, the idea is to really prevent that early dropout rate of 50% at the start of therapy, so that clients who do not have a good experience with a clinician early on and do not endure this problem, are being matched and receive treatment, proper treatment. We we do know that clients that have a bad experience might never return to treatment. And so this is it's really key to be able to address it at the start.

Client Experience And MVP Status

Dr Andrew Greenland 14:06

I hear you. So from a client perspective who's um registering on the platform, without having to give away too much because I know it's proprietary, what does that intake process look like? And what are they seeing? What's the output? Are they reading reading things? Are they seeing an avatar? What does it actually look like to a client?

Dr Ayelet Hirshfeld 14:24

So they are actually um asking questions and speak about what help they need, and they're interacting and speaking to the virtual therapist persona online, and so they can ask any question they want, they can ask about how the therapist will approach different um issues, how they interact, they can ask about their expertise. They basically ask any question they want and are able to speak to that by right now, it's by writing, and but are able to communicate with the virtual therapist persona.

Dr Andrew Greenland 15:07

And so far, do you have any kind of preliminary results of how this is working out?

Dr Ayelet Hirshfeld 15:13

So we are really only at the MVP stage, and we have not opened the product yet to a closed beta, which is what we are doing very soon, what we're gonna be doing very soon.

Dr Andrew Greenland 15:29

Got it. So you're kind of in beta, I guess, at the moment.

Dr Ayelet Hirshfeld 15:31

We're still in the development phase and we're almost done with the MVP.

Dr Andrew Greenland 15:37

And in terms of when this um does get rolled out, will this be the um therapist that will be funding themselves to be registered on the platform, or will it be the client

Funding Model And Rollout

Dr Andrew Greenland 15:47

paying for the service? I'm just trying to get someone to understand how this will get funded.

Dr Ayelet Hirshfeld 15:51

In terms of the go-to-market, it will be the burden will be on the therapist side on the system side, and whether it is an individual therapist or a bigger organization, not on the client side.

Dr Andrew Greenland 16:03

Okay, thank you. So you've got fairly deep roots in psychoanalysis and trauma work from your background. How do you see that integrating with innovation and tech going forward?

Dr Ayelet Hirshfeld 16:14

Well, the the way that you know, part of my role is not only as the founder of Psychology Inc., I also advise different organizations as a health consultant, as a trauma expert,

Where AI Helps And Where It Lags

Dr Ayelet Hirshfeld 16:28

and some organizations within the um area of AI and mental health, and and some not. And um I really see AI transforming mental health care in a very, very significant way. And I think that it's going to answer many different problems and be able to help with the problem of discontinuity of care, fragmentation between mental and physical health care, and the rising exposure to chronic stress and trauma from economic pressure, climate-related events, social instability, and which all increase demand whilst training limited resources. And I think that, you know, as a mental health clinician, um I see that trend of practitioners that are really adapting AI across mental health care. What we see mostly right now within AI is focused on reducing administrative burden, and things like transcription of sessions, automated note-taking, documentation support, billing, scheduling, which is really, really important to reduce clinician burnout, free time, and allow therapies to focus more on clinical work rather than paperwork. But what's relatively still undeveloped or underdeveloped is AI systems that address the relational and structural challenges, how people are matched to care, how continuity is supported, and how we prevent early disengagement. And this gap is where I see the next phase of meaningful innovation that needs to happen. And so, again, you know, when you look at it, matching is not really a clinical intervention. It's not a clinical intervention, it's not an administrative afterthought. And ethical AI should and does reduce friction and give clinicians time back. It does not replace their judgment, it does not replace their care. And so this, I think, answers a very big question. And I think that this is something that a lot of clinicians are worried about, and not only clinicians, and the question of can AI replace therapists? And I think that we need to change that question and look at how AI can strengthen, can really make the human care stronger and not replace. And so if we do it right, um it can really better care. And that's why we built ANA, which is the system of matching clients and therapists, and to be able to again not replace therapists, but better support care.

Dr Andrew Greenland 19:21

Thank you. So you talked a look about a lot about ethical AI and how that can benefit you know things like client and therapist matching and streamlining some of the operational

Ethical Risks And Safeguards

Dr Andrew Greenland 19:31

things behind the scenes. But let's talk a little bit about unethical AI because there's been a lot of stuff out in the media about how AI is being used to give therapy, which obviously that's something you probably don't approve of. But what are some of your concerns around AI in the more in the wider sense?

Dr Ayelet Hirshfeld 19:47

Yes, well, um we've seen some examples in the news of um um different uh platforms and where the safeguards were not yet there, I would say, I think that it is getting there. And clients that interact, not clients, but individuals interacting with these platforms, asking questions such as um how can I harm myself, and where these platforms would actually give them instructions on how to and will fail to flag um serious harm. And I think again that um, as you say, the question of ethical AI in these instances, um, AI can cause harm, but it's not so much um the system itself, it's making sure that there are safeguards in place that are being adapted. And the question of educating the public, educating them that AI is not a therapist, and the problem that we're facing, that individuals are going online and are using all of these platforms, whether it's ChatGPT, Gemini, for mental health, um, which is a phenomena that we cannot really prevent, right? Aside from educating the public and having safe guards in place and that will flag, for example, suicidality, homicidality, and really redirect you to the right resources and educate the individuals that this is not instead of a therapy, and they should reach out to a therapist. And so these are, I think that this is where um a lot of the work lies ahead. Um and yeah,

Adoption Barriers In Mental Health

Dr Ayelet Hirshfeld 21:35

this is this is an issue we're I think all facing.

Dr Andrew Greenland 21:40

So ethical ethical AI is clearly moving at pace, and obviously your platform is going to be doing huge amounts to improve the matching and the efficiency of a client therapist interaction. Is there anything in mental health care that you don't think is changing fast enough from your experience?

Dr Ayelet Hirshfeld 21:58

I I do think. That adapting AI to better mental health care and to be able to really address these different facets and that the mental health care industry is suffering from, and it's not adapted fast enough. I think that it requires a lot of education within organizations, and as we see AI being adapted in organizations outside of mental health care. And I do think that clinicians are worried of technology, and there are a lot of legitimate fears at the same time that it can really provide very, very significant insights and be used responsibly and ethically mental health care treatment, make sure that it's more precise and produce better outcomes.

Dr Andrew Greenland 23:06

Thank you. So obviously you're involved in a number of different things, clinical practice, the various platforms.

Wins, Impact, And Mission

Dr Andrew Greenland 23:12

And the platform is in development. What are the things that are going well from the point of view of developing this that you're sort of proud of, happy? Just curious to know.

Dr Ayelet Hirshfeld 23:24

Yeah, well, it's been a very fascinating journey. I think that it's a very significant addition, innovation that we're introducing. And that is a problem that is often overlooked. To be able to really focus on that and come up with a product that that answers, that fills that gap has been very fulfilling. And as a clinician, as a clinical psychologist, as a psychoanalyst, you do a lot of meaningful work, which I'll always keep doing, one-on-one, seeing clients, and which is a work that I love. But at the same time, your ability to make a big impact on society is more limited. And so I do engage in different things, such as being an editor of a psychoanalytic journal, being able to advise organizations, consult organizations to make a larger impact. And with this product, I really hope that the rich and the impact that I can help, that I can have to better improve mental health care and will be

Challenges: Compute And Scale

Dr Ayelet Hirshfeld 24:40

more significant, which is emission and that you know, sense of emission that I'm I'm on.

Dr Andrew Greenland 24:50

Thank you. And on the other side of the coin, what are some of the challenges in you know one getting a platform like this off the ground, but also the challenges of patient facial work and all the other things that you were involved with clinically?

Dr Ayelet Hirshfeld 25:04

Well, you know, it it does take a lot of compute power, and so there are significant financial costs associated with that. Um at the same time, um, you know, this is exactly where we are, and um where we are in terms of um launching the product, seeking funding to make sure that um we can continue development um at the pace and that um the pace that it requires. Um and the other question had to do with my other worlds or clinical practice.

Dr Andrew Greenland 25:40

Can you just in general, the things that the are the most challenging at the moment or frustrating in the work that you do? I'm just curious to know what that looks like from your perspective.

Dr Ayelet Hirshfeld 25:50

Well, that's an interesting question. I I would say that I always look at frustration or challenges as an opportunity for growth. And I always keep learning. And um I do um again, you know, the work is always challenging, it's always intellectually stimulating.

Trauma Work And Global Events

Dr Ayelet Hirshfeld 26:12

And so I would say that one thing that comes to mind that is very, very significant in my work is um events that that happened globally, such as um October 7th, for example, and being able to really help and support different organizations that are helping and providing support for individuals affected by these tragic events. And um I am a consultant to the National Institute of Neuropsychological Rehabilitation in Israel, and so that really allowed me to um to uh integrate um my knowledge in trauma and help support programs again that uh specifically help support um survivors of October 7th. And that is something that's meant very meaningful. Of course, it is um um it is a very I would say it is a very emotional topic as well, being Israeli myself. And so yeah, this is you know part of responding to to real events happening during real times and being there to support clinicians who are doing the work in the field.

Dr Andrew Greenland 27:45

Thank you.

Staying Grounded And Burnout Support

Dr Andrew Greenland 27:46

And I that leads me nicely onto my question about burnout. So, in a field with so much burnout and overwhelm, how do you stay grounded, so both professionally and personally?

Dr Ayelet Hirshfeld 27:57

Well, uh I um give this advice to a lot of clinicians to me to make sure that they take good care of themselves. And I practice a lot of different sports and yoga, uh, pilates. I think that's really, really important to take good care, good nutrition, make sure that you prioritize sleep so that you're able to be focused, sharp, and feel well yourself. And this is part of what I advise organizations as well, to make sure that the organization is healthy. Um, and I also consult to startups that do exactly that in the intersection of mental health and AI. Um, and um this is um something that you know um I cannot um say much about it, but I'm a um scientific advisor to uh Resis Labs, and they tackle exactly that problem of supporting clinician burnout with AI. Um and um this is again, this is something that I've been involved with across different organizations and such as as well as international SOS, providing support to clinicians in different areas who struggle with compassion, fatigue, and burnout, and being able to really share these uh insights with them to make sure that they better their health and their family's health.

Dr Andrew Greenland 29:28

Thank you. So, what's next? I guess the the platform being in development, that's your main baby at the moment, the thing that you're working on. But is there anything

What’s Next: Ventures And Growth

Dr Andrew Greenland 29:36

else on the cards going over the next six, 12 months that you're looking forward to or you're working on, or you're excited about?

Dr Ayelet Hirshfeld 29:42

Um yeah, I'm working and I I keep uh I would say um adding on to my uh portfolio different companies that I'm advising. And I uh joined as a partner in a venture. I'm I'm a partner at Unlock Pacific Ventures, a venture fear. And so I hope that I can also start supporting innovation. And because again, uh funding is a very significant component of being able to launch the products, to bring them to the public, to share the innovation and better the healthcare field as a whole. Um I'm very, very um excited about that as well.

Closing Thanks And Reflections

Dr Andrew Greenland 30:31

Thank you so much for joining me this afternoon. It's been a really incredible conversation. So thank you for pulling back the curtain on such a complex issue. I think it's really exciting to see a solution that isn't just plug in tech for tech's sake, but actually really respects the relationship between the therapist and the client. So thank you for sharing. Thank you for telling us about your practice, and I wish you obviously all the best with the development of the platform and your endeavours going forward.

Dr Ayelet Hirshfeld 30:53

Thank you so much. Thank you for having me here.