Leveraging AI to Create Genius Healthcare Experiences
Healthcare consumers expect seamless, integrated experiences - but the fragmented U.S. healthcare system makes this challenging. In this webinar, learn how innovative AI solutions are bridging this gap and creating meaningful experiences that deliver real value.
Speaker: Tiffany Mura
When we look at experiences, we understand that consumers are entering into them with jobs to be done. They are hiring a company or a service to help them accomplish this job. And we at Stone Mantle consider that jobs fall into one of four categories. They're either functional, where the consumer's trying to accomplish a task, emotional, where they're trying to feel deeper about a particular moment, social, where they're trying to help connect and get support from others, and aspirational, where they're trying to change something about themselves, whether it be physically or mentally.
So if you think about that in the healthcare context, a functional job might be booking a doctor's appointment or scheduling a procedure. An emotional job would be helping them get support if they were receiving a difficult diagnosis. A social job might be a consumer looking to find others who suffer from a disease state that they're experiencing to find support in that community of others. And aspirational would be a job where the consumer is trying to improve their health in some aspect. Perhaps it might be losing a large amount of weight or starting up an exercise regime. The challenge we've got in the healthcare space is that healthcare is often unable to get the entire job done by a single entity.
And that problem exists for a variety of reasons. One, it's the fundamental nature of the US healthcare system in that the system is not a single payer system that is fully integrated where the information is flowing seamlessly through the system, different companies and different organizations own different pieces of the journey. And also there's some regulatory limitations around this. So how do we overcome this? How do we help healthcare consumers get the whole job done? And one of the reasons this question is so important is because consumers now come into their healthcare experiences with an entirely different set of expectations than they previously had.
Historically, they were willing to accept this situation in the healthcare system that no single entity could really help them complete a job end-to-end or complete a job systemically. But with the proliferation of digital solutions and technology solutions elsewhere in their lives, they're now saying, well, I want that same sort of experience within healthcare.
So we at Stone Mantle work with our clients to help make the experiences better, help overcome these challenges, both for regulatory and systemic situation, and really leverage technology and data design and tools in order to create better experiences. And one of the ways we measure the success of these tools is on a spectrum of what we call kind of stupid to genius.
If you want to look at this as a linear hierarchical set of criteria, the idea is taking a look at whether or not the technology and the data and the experience are being designed to do more than the consumer could do for themselves. So if you're thinking about at the lowest level, a stupid or dumb experience isn't helping them get the job accomplished at all, or it's making it harder for them to get the experience done and get the job done.
Moving up the ladder, the job might be a little bit easier to get done with the technology or only addresses a tiny part of it. What we really wanna help move companies into is creating experiences that range from smart to genius. And the idea is that in a smart experience, you're using technology and data and experience design to deliver intelligence to get the healthcare job done in a more complete and more systemic way.
But we really can take that out into the stratosphere when we say, well, how do we really leverage the best of technology and the best of data to augment the skills that a consumer has and make their abilities almost into superpowers so that they can do so much more with the technology, the data, and the experience than they ever could do on their own?
So if we look at an example, take a migraine application, for example, you might have a migraine application that is given to a consumer to track their migraines. And in the lowest form of design, this sort of dumb experience, it's only doing part of the job. It's allowing them to know information about migraines, what might be a trigger, but it's not allowing them to do anything that's uniquely relevant to their own situation. So you move it up into the smart experience by giving it a systemic view of their lives and a personalization element where it allows them to track their individual migraine information, their trigger tracking. It starts to use that information to learn about their own disease progression. And it provides them with tools to come into a doctor's visit and be better prepared for a discussion because they've got a symptom log, a trigger log, and they can have a more intelligent discussion with their doctor.
But where we really elevate the experience is when we turn it into genius and we have it do multiple jobs for their migraine situation and look at it in a systemic view of their whole health, not just their health related to their migraines. So in that case, that type of experience would take all of that individual information that it's learning and not only track it, but then use it to predict potential attacks. And it would look at the migraines in relation to what else they might be doing. Perhaps it's taking in information from their smartwatch on how many steps they're getting every day or how much sleep they're getting and starting to integrate that with the migraine tracking and using that information to both predict and help them stop triggers and then attacks. It would also take into account what other diseases they might be suffering from and have those treatment plans worked into that system. And like I mentioned, it would be connected not only with their smart devices they're using personally, but also with the systemic technology like their doctor's EMR system.
At Stone Mantle, we've been doing a study now for four years where we take a look at these smart and genius experiences. And one of the things we've done in order to have more intelligent conversations with consumers and help them grasp this concept of genius experiences, we've quantified eight different healthcare superpowers that consumers want.
The first is memory champion. This idea that they don't, as a consumer, have to worry about remembering to take their medication or their treatments, and it allows them to stay on top of their health by not only tracking what they're taking, but proactively reminding them when to take it and ensuring that they're not taking things that are interacting with each other, as well as giving them proactive reminders of when it's time to check in with their doctors or time to get labs done, and really augments as kind of a second brain for them.
The second one is mind shifting. And in this one, the idea is that it really allows them to have support in shifting their mindset so that they can move from mode to mode to a mode that might not be as supportive of their healthcare goals to a mode that is. So for example, if they're trying to reduce their blood pressure, it helps them go from stress to relaxed, or it might help them go from unmotivated to motivated. And this is a type of tool that would help them throughout the day. And it would proactively understand kind of where their mindset was and help them move back into the healthier mindset that they are trying to accomplish.
The third one is health foresight. And this really helps their doctors foresee future health issues that they might encounter, whether it's a chronic disease like high blood pressure, high cholesterol to a more concerning condition like cancer and help stop those conditions before they start. Whether it's using biomarker data or it's using family history data, the idea is that it'll be predictive and let them know what conditions they might develop and tell them the steps they can take to prevent it.
The fourth one is lifestyle configuration. This is really kind of the coach in the system, giving them guidance and assistance on how to make adjustments to their lifestyle to achieve the healthcare goals that they are trying to achieve. So it might be things like telling them what to eat and when to eat, how much sleep they should get, how much water they should drink, how much exercise they should do. It also schedules rest times for them, plans meals, generates shopping lists. So it really acts as that coach and that support system to make it a lot easier to do all the things in their daily life that are necessary to improve their health and accomplish that job of getting into better shape or lowering their blood pressure, whatever their particular goal might be.
The fifth one is hyper-understanding. This is really where the consumer feels more heard and understood. And the information, the data that they know about themselves is clearly communicated and it's put together in an integrated way. We all know we can go on to the EMR now and see all of our healthcare data, but it's not categorized or shared in a way that's really useful and actionable. And the idea of this hyper-understanding is really taking the test results, the treatment plan, the diagnoses, the symptoms, medications, and putting it together in a really integrated view and making it actionable in a way that makes the person feel more cared for and more confident in their ability to move forward whatever the treatment plan is.
The next one is body intuition. This is kind of like a car tool, diagnostic tool, and the biometrics of the person using it would be linked to smart devices that identify information that tracks their health, just like a car diagnostic tool would. So it reminds them of things like standard maintenance checkups, you know, as a time to go in for a physical or time to go in for lab readings. And it also would identify when there might be an emergency situation or a more urgent situation, like when the check engine light pops on, to indicate that they have something going on that might need more medical attention.
Just right motivation is the next one. And this is a way to help the person accomplish their health goals by proactively sending them activities that match their interests and their healthcare objectives so that they can participate in more healthy activities and stay motivated.
Lastly is information sharing. This is where the information is all aggregated and shared, not only for the patient themselves, but also for their doctors, because that is still quite siloed, depending on whether you're seeing doctors that are within or outside of a particular healthcare system.
And this is the idea of bringing that all together in one spot. So we work together with consumers to develop these, to quantify the types of superpowers they might be looking to achieve through experiences that combine smart digital data design and technology.
How AI Delivers Genius Healthcare Experiences
What are AI agents?
Programs designed to make decisions and complete tasks on their own.
Autonomy
Reactivity / Proactivity
Learning and Adaptability
Context Awareness
Goal-Driven Behavior
Scalability
Interoperability
Speaker: Shai Rozen
So the question at hand is how do we use AI and how do we use advanced technology to deliver these genius healthcare experiences? So one of the ways is through this AI agents, as we call them.
So before we dive deeper into that, I just want to take a moment and quickly do a definition. What are AI agents exactly? So AI agents are programs of pieces of software, pieces of code are designed to make decisions and complete tasks on their own. There are a couple of characteristics that are sort of fundamental and very, there's definitely more, but these are some of the key characteristics of agents.
First of all is autonomy. These AI agents can work in a fully autonomous mode. We've seen in some organizations that in many cases, that's still too much to put it in full auto mode. But that's one of the capabilities that they have, or we can have them sort of set to a more human mediated level of autonomy.
Then is reactivity and productivity. Reactivity speaks about the capability to these agents of reacting to some sort of change in variables, either through, let's say, some person's wearable device data or some sort of API integration or something that triggers certain behavior or certain action from the AI or even sort of all the way to the proactive side of things of them being capable of saying, OK, this needs to be done now and proactively taking some sort of action of completing some sort of task.
We all know sort of the nature of artificial intelligence is learning and improvement. We're not all the way to self-improvement, but we can rapidly improve and rapidly learn new behaviors and most importantly, correct things that are not exactly right, and adapt to the circumstances.
Now, the adaptability piece, it's extremely interesting because it's not only adaptability on the long-term, it's sort of real-time adaptability to the situation. So changing communication modes, changing the way something is explained, and even self-correcting some things when, I don't know if you've had this experience, where let's say you're working with a tool like ChatGPT, and then you say, hey, you made a mistake. It's all self-correct. You say, oh, yeah, you're right. I'm sorry. Here's something. So that adaptability is key.
Context awareness has to do with that reactivity and productivity capability. It has to do with being aware of what's happening to the person in front of them. And this can be location. It can be mood. It can be symptoms. It can be pretty much everything that gives the AI or gives the agent a better understanding of the circumstances in which the actions that it's doing are taking place.
Usually, agents are goal-driven. They have a set of tasks, a set of use cases. There are more or less sophisticated agents. Some are very, very specific. They do one thing very well. Some are a bit more complex and comprehensive, and they can do multiple things. But usually, everything they do is very oriented to specific tasks or specific goals. Scalability is key in every software technology functionality, nothing to say much beyond, obviously, if one agent works, there's no limit to how many replicas of that agent or how many interactions it can be having at any moment in time. And one of the things that's super interesting, is sort of this idea of interoperability. So being able to not only draw data from different sources, but also take action on third-party services in a similar way that a human would do without the need of very complex integrations. I think one of the most common things that I hear from organizations is that scary look in their faces when we're trying to think about, okay, we're going to implement a new software system, and then the idea of how do we make it work in the massive ecosystem of different databases, different sources of data, different sources of knowledge.
One of the amazing things about these agents is that they can very rapidly and very simply connect in multiple ways to different systems because they can. It doesn't mean it's always like that, but it can interact with a system in the same way that a human would.
What Are Some Characteristics of Agents That Can Give Superpowers to Their Users?
Infinite long-term memory (memory champion)
Personalized and adaptive communication style (lifestyle configuration, mind shifting)
Need a drill sergeant today or some chicken soup from Grandma?
Proactive outreach and context awareness (just right motivation, hyper understanding, body intuition)
Stepped into a restaurant? Get a note about what to eat.
Forgot something? Get a reminder.
Achieved a goal? Get a pat on the back.
Deep data “understanding” and access
Communication with 3rd party services (information sharing)
Share relevant information with providers or caretakers.
Create support tickets, schedule consultations, etc.
Order groceries, fill prescriptions.
Predictive analytics (health foresight)