Can AI Fix Healthcare With Trust

Rivka Allouche (00:01.532)
Today I'm honored to welcome Nadav Shimoni in the Care Culture Talks podcast. Nadav, you are an early stage investor at A-Squared Ventures, focused on digital health. You partner with founders who are rethinking how healthcare is delivered and experienced. Your interests are at the intersection of technology and patient-centered care.

We are very happy to have you on the podcast. And so according to you over the past year, what changed most in the AI technology in the healthcare sector?

Nadav (01:03.212)
Well, thank you first of all, it's great being here. I'm very excited to be an investor and working with No Barrier. For your question, well, a lot has changed. If I need to pick one thing I would say right now, perhaps for the first time, there is an opportunity for software, for AI to really tackle service-centered issues. Putting it differently, when everyone speak about the five trillion healthcare expenditures in the US, most of it is services. And only, not so while ago the real value proposition of software was tapping into IT budgets which is a big number but it's a small fraction compared to the services budgets of US healthcare. And now because what we're seeing with AI We can actually tackle or dip into these services budgets and I think No Barrier is a perfect example for that.

Rivka Allouche (02:17.806)
There is still a debate between AI and human. It's really here. When We talk to people and on the other hand, we have a practice manager who tells us like we've been betting on technology for a while and it's like the like a natural step for us to dive in. So what would be your take on that?

Nadav (02:48.546)
I think healthcare revolves on trust, either between patients and physicians or between physicians and physicians and even between physicians and insurers.

Everything is hinged on creating trust and it's very hard to create trust just with software, just with AI. And by the way, for this conversation when I'm saying software, I'm meaning AI and vice versa. For me, it's the same thing right now. So to create this element of trust, I really think a human element is crucial.

The key question for me as an investor is really trying to figure out how you can make that component as efficient as possible, how you can do essentially more with less and automate more and more as the company grows. So long story short, I think the best solution is actually a solution that combines human and software or AI and is able to be become more more efficient on the human component as it grows.

Rivka Allouche (04:00.951)
And now for being a little bit more specific, what would be the difference between having a more generic software or more generic tool like Google Translate or other general tools compared to another tool very focused on healthcare?

Nadav (04:22.702)
Yeah, I think for me it's the 80-20 question and actually in healthcare it's not about the 80%, it's about the 20%. Building a product right now... like generally building a product right now Having 80 % off something is relatively easy. I mean I can while we speak, you know vibe code it or code code it or call it whatever you want. But what actually matters in areas like health care where you have lots of Concerns, you know for privacy, for security, for workflow disruptions... Is these 20 % that comes after that, which are, know, the integration, deployment, ease of use, and so on and so on. So that very intimate acquaintance with the domain you're building at, that specific toolbox that is required to do that. These are very important elements right now. So, you know, obviously, and objectively, I'm speaking from a position of you know, being a dedicated healthcare investor. But I do think having healthcare focus is very important. There are some areas that could be tackled by a generalist or a general products. These are the kind of the lower threshold where most are lower. Think about scheduling tools, I think is a great example. But the deeper you go into the workflow, I think there is a very important requirement for specific approach and healthcare focused approach.

Rivka Allouche (06:03.705)
I agree. We talk a lot about the pain points and what the clinicians feel and also the staff, so the people who are actually in the front line. Burnout is still one of the biggest challenge in healthcare and do you see that digital health and all the tools around digital health will help?

Nadav (06:34.274)
No, For sure, as a clinician and as a former operator in health system. First and foremost, I think that's the hope, right? But there is a big gap so far in many areas between that hope and reality. And I think that part of this gap is a consequence of perhaps some naivety of innovators trying to innovate from outside without fully appreciating the difficulty of the day-to-day clinicians are experiencing and the complexity of what We call clinical workflow.

I have experimented with different solutions based both as a clinician and as an operator and now as an investor. And I think, yes, there is a very tangible opportunity to do that, but it is requiring, and it goes back to your last question, lots of prior thinking and lots of good design together with the staff to make sure it is actually happening. and Studying it and refining it and maybe from the beginning it won't happen but I think We can get there in many places but it requires a lot of dedicated effort.

Rivka Allouche (08:04.343)
It's a good answer and it directly leads me to my next question. So right now, from your experience, what do health systems really expect from a startup that comes with a new solutions in terms of integration, adoption, workflow, also costs?

Nadav (08:22.786)
Sometimes there was a misperception for, from companies that even if somebody's giving you like, hey, I'm giving you my product for free. Even if I'm giving you as a health system, my product for free, it's actually not free because, the burden of integration and change management is actually, in time and perhaps even, bothering the staff type of resources. It costs the health system quite a lot. So even free is not really free. And going back to your question on that end. I think right now, there are different boxes a company has to check in order to make sure they are compelling enough. One is a very clear and objective financial ROI.

Either there is a clear budget item for what they are solving and they are reducing that budget to a meaningful degree or there is a very clear additional revenue they are contributing by working with their partner. And you need to have data to support that claim. You cannot just hand wave your way in. So that's one. Two. Frictions less integration.

The more you ask from the IT and even if it's a small ask, I think the chances you will start working with your potential customer is very low. So there are great and innovative ways to make integration frictionless today. But it's a lot about thinking about your product, perhaps having an initial offering, and then expanding it and so forth. But the first step in should be as smooth as possible.

Rivka Allouche (10:23.649)
And so I like when you mention about budgets and costs. I recently came across a paper from Katy Haynes from the California Health Care Foundation. And so she was also herself quoting Jeffrey Miller, one director of health of another institution. And so he was saying that a cost reduction is nice, but at the end you don't want, you know, to they deteriorate the quality. How do you get to this balance?

Nadav (11:09.538)
You know, I think that's the silver lining of innovating in healthcare. You have perhaps as opposed to other industries or maybe in a more tangible manner, you need to address both quality and cost or like both the clinical and the financial elements altogether and separately. And even on top of that in healthcare, you might end up with three or four different stakeholders that are relevant to the buying decision. Some are because of the usage, because of the budget and so on and so on. And you need to make sure all these stakeholders are satisfied. So 100 % agree, you cannot hurt the clinical outcomes even if you're improving the financial ones.

Nadav (12:08.512)
I mean the magic happens when you are able to address all of these altogether. And quite frankly, I think No Barrier is a great example. Like if you can offer a product with a fraction of the cost that your customers are willing to pay right now, but on top of that you also provide a much better experience.

Then, that's magic, right? And you see the consequence of that magic. So that's the challenge, but also the opportunity. Because when it happens, it's like finding a needle in a haystack.

Rivka Allouche (12:48.601)
I can share like a little story. We had in a podcast, a guest who told us that actually one of his nurse told him that the experience with AI was outstanding comparing to a human interpreter. Whereas we are always trying to find the right balance to opt for a hybrid model. But actually in some cases it was actually in peds in which you really want to have the small talks to be interpreted. "You have such a cute child" and it really helped the conversation between the nurse and the parents. So it was actually very nice and surprising to hear.

Nadav (13:34.414)
When it works, again, it's like magic, sadly in many cases it does not. And for me, that's my task as an investor, trying to predict where it might actually work.

Rivka Allouche (13:51.512)
And in terms of prediction, for you more into the language access, what would be the picture of language access and the integrations of the technology in the next 10-20 years?

Nadav (14:10.488)
Well, I would actually provide a shorter timeframe.

You know, I don't think it would be an exaggeration to say that language services right now are experiencing some sort of like a Kodak moment. It's like it's becoming digital, no matter if you want it or not. And this notion of I'm picking an iPad and I'm trying to connect with somebody based on the other side of the world to provide language services and my connection is breaking up and perhaps this individual is not the best fit for this type of translation.

This is not 2026. And I would expect some of the traditional language service providers will try to keep their business because that's their business and they will end like that manual photography businesses a while ago. And some others will innovate their way. But these organizations are now experiencing the core of the innovators dilemma, right? They need to change their main business their main business right now.

Nadav (15:25.95)
So I think there is a real opportunity to see a change in the foreseeable future. I think we're starting to see it with companies like No Barrier. And I would expect these services of language services will go deeper and deeper because it's a barrier to treatment, right? It's a crucial element to provide best-of-breed clinical care.

As a physician really understanding the history of your patients and really understanding what's going on with your patient, that's the foundation to any care you can provide and if you get this wrong it can lead to so many bad things so I perceive language services as a crucial element in providing care.

Rivka Allouche (16:20.695)
This leads to another question more about regulation and policy. I see a lots attempts of writing papers and to see how to manage AI and technology. We can say that people are afraid or maybe they want to have more control about the possible risks in a healthcare organization. But I've also heard other clinicians saying like at the point of care when you have an urgent situation you just need to provide. So what would be the best practice or how to handle policy and regulation with AI technology in language access?

Nadav (17:08.92)
Well, I think it's a big question and many people are trying to tackle it and lots of smart people are trying to offer solutions. You know, I think like with every big change in tech, actually it's not a tech question. It reminds me of this notion or a story I've heard that when cars started, they've limited car speed to be at the speed of horses to make sure the cars are not interfering with the horses and so on and so on. You can find different comparables or like...

Never mind, I'll stop with the history but I think right now it's actually a question of what is the risk We are willing to tolerate as a society or what does it mean to have AI going or making mistakes because by all means We all make mistakes as people, right? And human interpreters make mistakes. It's not flawless. So the question is

Do We compare AI to humans? we perhaps put a higher threshold because we expect it to be more efficient? Or who is responsible when the AI is wrong?

Essentially, and I think there could be very good answers to all these questions. And to your point perhaps where we see things progress in the fastest way would be in areas where we don't have enough supply of human resources. So we will have to defer to AI and perhaps in other areas it will be slower. But I think by the end of the day, it is a very addressable question and I would expect we'll have more clarity on it sooner rather than later actually.

Rivka Allouche (19:13.337)
I like your example about the car and the horse. We are maybe just learning and like everyone wants to see where's the responsibility, where's the risk. But I really hope in the sense of language access and equity to care that it won't block the efforts of what the technology can bring.

Nadav (19:37.302)
100%. You need to balance between risk and reward But again humans also make mistakes, so it's just the way it is.

Rivka Allouche (19:49.595)
I agree. I think we covered a lot in these almost 20 minutes. Are there other insights you would like to share with us and maybe our audience would be happy to hear?

Nadav (20:03.406)
I actually think we did cover a lot. It was a pleasure speaking with you and I appreciate the opportunity.

Rivka Allouche (20:11.301)
Thank you very much for all these answers and I hope to have you again soon on the podcast.

Nadav (20:18.03)
Sure thing, my pleasure, thank you.

Rivka Allouche (20:19.782)
Thank you.