🦄 Making great presentations more accessible.
This project enhances multilingual accessibility and discoverability while preserving the original content. Detailed transcriptions and keyframes capture the nuances and technical insights that convey the full value of each session.
Note: A comprehensive list of re:Invent 2025 transcribed articles is available in this Spreadsheet!
Overview
📖 AWS re:Invent 2025 - Next level patient experiences with Amazon Connect powered by Kyndryl (AIM105)
In this video, Edward McIlheran from Kyndryl and Sunil Menon from AWS demonstrate how AI-powered patient engagement transforms healthcare contact centers using Amazon Connect. They showcase a demo featuring conversational AI with Amazon Lex and Polly that achieves 40% call deflection through natural language self-service for appointment scheduling, identity verification, and care coordination. Key benefits include omnichannel communication, personalized care through EHR integration, and streamlined administrative processes. Jupiter Medical Center's case increased radiology equipment utilization from 75% to 95% using outbound campaigns. A federal government implementation achieved 30% call deflection in the first month. The speakers emphasize Amazon Connect's modularity, consumption-based pricing, and crawl-walk-run adoption approach for modernizing patient engagement across the entire care journey.
; This article is entirely auto-generated while preserving the original presentation content as much as possible. Please note that there may be typos or inaccuracies.
Main Part
Introduction: Transforming Patient Engagement Through AI-Powered Contact Centers
Thank you everyone for coming out today and spending a few minutes with Sunil and myself. My name is Edward McIlheran. I'm with Kyndryl, and I'm a Director, Consult Partner with our digital workplace services practice. I'm focused primarily on contact center modernization across all of our different verticals, including healthcare, state and local government, financial services, and many more. Hi, I'm Sunil Menon. I'm a Senior Business Development Leader for Customer Experience Engagement in US healthcare at AWS.
There's a QR code that we'll have this slide up at the very end. If you want to hang out afterwards, we'll be here for the next half hour to discuss any questions that you may have stemming from today's presentation. The short of our presentation today is how can we help you engage your patients during the entire episode of care and beyond, leveraging the power of AI. I get it that AI is the number one thing happening in our industry. Everybody's looking to AI to understand how we use it to solve our key use cases and really drive home a lot of benefits for our organization.
So today, we're going to focus on three themes that I'll talk about here shortly, and you'll see that throughout the deck as well, showing how this applies to patient engagement and what are the opportunities to leverage AI in the contact center space. Number one is the consumerization of IT. We're seeing a lot of chatbot experiences that ChatGPT and similar platforms have made accessible to everyone. It's really revolutionizing the way that you can interact with data. Now consumers have the preference of having this from the services they consume, including healthcare, so it really opens up the opportunity to reimagine front-end engagements with your patients.
Number two is eliminating long hold times and inefficient activity. With a combination of a shortening workforce and also more call volumes coming into the healthcare space, this really is a recipe for a potentially bad experience. So how do we leverage AI to enable self-service capabilities and drive down call volumes while at the same time giving your caregivers better tools with AI that allow them to service patients faster? Number three is reducing the complexity around how care is delivered. In a modern healthcare scenario, you may have follow-up appointments, you may have to coordinate different care specialists, and sometimes you just don't know where to go to get help when you need it. We can leverage AI to look at this from a holistic view and be a shepherd, guiding patients in a smart way to make sure that they're following along and we can avoid scenarios like non-adherence and readmission.
Sunil, any thoughts there? I'll dive into how this patient journey unfolds on the next slide. If you really think about the patient journey, whether it's for a short episode of care like going for a cold or a longer therapy like a cancer therapy, there are some distinct steps. There is an intake process where patients communicate what they're going through. They need to get their appointments. Then there is the actual visit part, which can be virtual or in person, and then there's post-care follow-up and then there is nurturing campaigns and continuous engagement with proactive communications.
When you think about this on the front end, the point where patients come in, that ingress point, that's where you can collect a lot of information and provide that high bar customer experience that they deserve. This can be done by having natural language AI powered by Lex as well as Amazon Polly to try and provide self-service opportunities. Where is parking? What time are the hours of operation? I need to set an appointment. These things can be managed automatically. If you can deflect 40 percent of the volume, as many customers have achieved, what that does is it takes the load off your engagement platforms, and your agents don't have to provide answers to routine interactions. They can focus their energy and effort on the high complexity, high value interactions.
Amazon Connect: A Cloud-Based Contact Center Platform Built for Healthcare
Let's actually see a demo, and this is an immersive demo, so I would encourage you to listen to it carefully. There are a lot of things going on and you can see what that next level engagement looks like. Before we jump into the demo real quick, for those in the audience who are not familiar with Amazon Connect, this is a full-blown contact center as a service platform that was built for Amazon by Amazon.
It became commercially available in 2017 and has grown very quickly to become a leader in the contact center space as defined by analysts like Gartner. It's an entire end-to-end application that serves multiple stakeholders with different needs and capabilities.
From a patient experience perspective, you have the full omnichannel opportunities to integrate calling through voice, chat, SMS, video, in-app, and web calling. You can also integrate agentic frameworks or agentic chatbots with conversational AI. If you need to elevate a call to an agent, agents have unified workspaces where you can integrate all of your applications that are necessary to reduce contact switching, as well as your knowledge library that allows you to get the next action in order to solve the inquiry faster.
From a supervisor perspective, you have all the reporting options that you would need, whether it's out of the box or customized. There are also many tools available to supervisors that reduce a lot of the time that goes into quality assurance for your agents. This really helps with getting them more efficient, training them faster, and improving them faster, but really improving that time to value for somebody new to the organization.
From an administrator standpoint, there's a simple, intuitive GUI when it comes to designing your contact center and making changes on the fly. Finally, there are modern APIs and SDKs. What you'll see in our demo is that you can leverage these APIs and SDKs to really curate the experiences that you're looking to create for your customers.
The two key advantages that we see with Amazon Connect in the field are the modularity and the consumption-based model. The modularity means that you don't have to adopt the entire platform all upfront. There's no upfront investment and there's no licensing tiers that you have to worry about. You truly pick the services that align to your organization's needs and you build those.
The consumption model allows you to pay for what you truly use. With proper financial operations practices, you can really achieve high ROIs when adopting Amazon Connect. With that, we're going to walk you through a quick demo on what this new experience could look like.
Demonstration and Implementation: Real-World Applications of Conversational AI in Patient Care
Thank you for calling AnyHealth. I'm your virtual assistant. Our system recognizes this number. Am I speaking to Juan? Yes, this is Juan. Hi Juan, let me quickly verify your identity before we go further. I've sent a one-time code to your registered phone number. Can you please read it to me? Sure, it's 547-099. Thanks. Could you also confirm your date of birth? August 1st, 1984. Thank you for verifying, Juan. Our records show you're a designated caregiver. Are you calling today about Terry Whitlock? Yes, I am.
OK, I see your mom was recently discharged from our hospital on August 15th. Doctor Williams, as part of her discharge care plan, recommended an EKG follow-up appointment. Would you like to schedule that appointment now? I can help you find a convenient time. That would be great. Based on your previous scheduling preferences for Terry's appointments, I see you typically choose Friday afternoons. I have the following slots available with Doctor Smith on Friday, August 22nd at 1:00 p.m., 2:30 p.m., or 3:45 p.m. at the AnyHealth Maine medical offices. Don't worry about writing this down. I'll send all the details, including the address, in your appointment confirmation email. Would any of these times work for you? I can also check other days if you prefer. 1 p.m. sounds good.
Perfect. I'll schedule Terry's EKG appointment with Doctor Smith for Friday, August 22nd at 1 p.m. at the AnyHealth Maine medical offices. And Juan, since this is a new specialist visit for Terry, would you like me to quickly check if the existing referral from Doctor Williams fully covers this consultation or if any additional authorization might be needed from Terry's primary care physician, Doctor Chen? Sometimes different insurance plans have specific requirements. That's a good point. Yes, if you could check. Sure, just one moment. OK, I've reviewed Terry's current authorization status. It appears the referral from Doctor Williams is comprehensive and covers this initial consultation with Doctor Smith, so no further action should be needed from your end regarding authorization for this visit. OK, great. For this appointment, is there anything specific that Terry needs to do to prepare? That's a thoughtful question, Juan. Let me check the standard preparation guidelines for a new patient cardiology consultation with Doctor Smith. Terry should continue taking all her current medications unless specifically instructed otherwise. Wear comfortable clothing. Bring a list of her current medications and not eat or drink anything except water for 4 hours before the appointment.
That's a lot of instructions. Can you send these to me somehow? Don't worry, I'll include all these preparation guidelines in your appointment confirmation email for easy reference. Would you also like me to text this to you? Yes, please. The instructions have been texted to you. Additionally, you'll receive appointment reminders as we get closer to August 22nd. Is there anything else I can help you with today? No, that's all. Thank you for calling Any Health. Have a great day.
I hope everybody was able to hear it. If not, you could read some of the subtexts that were displayed. This demo really shows how you can take an average interaction and transform that experience by leveraging all the capabilities that come out of the box with Amazon Connect. When you're thinking about it, you're not looking at it from the lens of a contact center. You're truly looking at it as a patient engagement platform.
For the interest of time, I won't be able to go through all of these, but these are a lot of the common themes and benefits we're seeing with customers adopting this type of strategy, and you could see quite a few of these in that demo alone. Enhancing patient member engagement through omnichannel communication is the first benefit. The patient called in through a voice channel, but they spoke with a conversational AI, and they also interacted over text message. It's a seamless unified omnichannel experience that the customer enabled.
You're leveraging AI and automation for personalized care. The conversational AI in the background was interacting with key systems like your EHR to understand who they were, why they were calling, what they could be calling about, but also hook into other systems for prescription refills or appointment scheduling. Being able to leverage that creates a contextual but seamless experience for them. Finally, there's streamlining administrative and operational processes. These contact center platforms like Amazon Connect have sophisticated IVRs where you can embed a lot of your simple tasks that take up a lot of time for your caregivers, such as bill payment or rescheduling appointments. You can streamline all of this in the IVR to provide a better experience for that patient calling in, but also give time back to your caregivers.
This is really the strategy of how we're seeing customers adopt, taking it from a contact center but thinking about it from an engagement platform and really rethinking those new experiences. Walk us through on how you would get started in implementing this strategy. The big challenge customers face is that the patient engagement journey is very complicated. There are many parts of the organization interacting with the patients. Where do customers really start? The answer is they can start anywhere. Amazon Connect is engineered for the future to accommodate all kinds of AI, agentic AI, conversational AI, predictive AI, and generative AI. The important thing is that customers can choose the extent and pace at which they bring AI into their organization.
If you look at this slide, the top half shows the fastest way to modernize your patient engagement journeys. On the top left, you'll see that you can improve the IVR like you experienced in the demo. It was providing self-service opportunities with natural language powered by Lex, Nova Sonic, and agentic workflows. I would highly encourage you to attend the Thursday session by University of California San Diego where they talk about how they were able to leverage Amazon Connect and agentic workflows to improve self-service opportunities, have patient verification, improve first call resolution, and reduce average handle time.
The right-hand side of this slide shows outbound campaigns. There's a need to proactively reach out to customers. An excellent example is Jupiter Medical Center. They had a peculiar problem. They have expensive radiology equipment that was used only about 75 percent of the time, yet they had a lot of open referrals for radiology. They used Amazon Connect outbound campaigns to proactively reach out to those patients, and they were able to convert a lot of them into actual appointments. The capacity utilization of this expensive equipment went up from 75 percent to 95 percent. Think about what that means. Number one, more revenue. Number two, the productivity of those expensive technicians is higher. And most importantly, patients get the imaging that they need so that they can have the diagnosis and the treatment that they deserve and need. So it's a trifecta of benefits.
Success Stories and Strategic Roadmap: Partnering with Kyndryl for Healthcare Modernization
The bottom half of this slide shows traditional contact centers. One hidden opportunity is in the clinics and departments. You think of any healthcare system, the departments, ambulatory centers and clinics. A lot of conversations happen between the patients and those clinics directly, and yet there are these old style phones that are not AI/ML powered devices. You can deploy Amazon Connect there, take advantage of the agentic workflows and all of the goodness, and the best part is because it's a consumption-based model, it's not a license-based model.
Even in those low volume locations overall, individually they're low volume, but together they contribute a high volume of interactions. You can deploy Amazon Connect. Ed, do you want to take it next and explain how Kyndryl has deployed Amazon Connect?
Yeah, so I chose this scenario because it was something that we deployed earlier in the year and it really just kind of speaks to everything that Sunil and I are talking about, right? You can adopt at your own pace, but really start with the workflows or start with the needs that matter most to your organization. When a federal government COVID era program wound down around safety net benefit programs, all of the constituents of the state had to re-enroll at the state level, and the infrastructure that was in place was not able to really hold up to the volumes that were coming in.
Undoubtedly what led next was bad experiences. People were on hold for 2+ hours, sometimes more than 4 hours. They weren't happy. Agents were burning out very quickly, so the agency knew they had to make a change and really figure out how to leverage technology to solve a lot of the issues that they were facing. And so we helped design a new system that spoke to a lot of the self-service needs that they were requiring.
This was around checking the status of entitlements, how much money is on a card, when the next benefit is coming through, whether there are other benefits that someone may be entitled to, and many more other use cases. What it led to initially in the first month of deploying this was 30% call deflection, which was amazing. There was also a self-service capability, so that meant when the hours of this department or the call center was closed, people could still go in there and ask critical questions, especially around how much entitlements may be on their cards.
It was a big win. The NPS scores were starting to improve. The agents were much happier and because of the amount of data that they were able to collect in the IVR about the caller, the agents were able to quickly reduce how much time it took to handle a lot of the inquiries that were coming through. In a great partnership with AWS, they've been fantastic in helping support us along that journey along with the client as well.
So Sunil, let's take us home. Keys to success. The first basic thing is making the decision to move to a cloud-based patient engagement solution like Amazon Connect, because if you don't move to the cloud, you cannot take advantage of all the data, AI, the continuous interactions that your patients are having, being able to personalize the experiences to them and understand their context. So that's the first step.
The second step is crawl, walk, run. It can be pretty overwhelming if you try to plan this whole journey and try to massively change your system. The trick is to find those opportunities like I explained, those IVR modernization as well as the outbound campaigns are good places to start because you'll get the ROI, you'll have savings and you can then pour those savings into modernizing other parts of your system.
The third part is choosing the right partner. Kyndryl is a great partner with a lot of experience in healthcare and Amazon Connect. Getting your people in the journey is very important because they have institutional knowledge. They know the inside of your organization. And when they work with the partner and AWS technology, they are going to uplevel their skills and be ready for the future so that they can deploy agentic workflows not only on the front end, but then take it downstream through all the organizational processes, including clinical workflows.
We're right out of time, so I'll quickly go through these last ones, but Kyndryl, we've invested heavily in the healthcare space. We have over 13,000 practitioners whether they're supporting our existing clients or involved in our consulting organization. We've invested heavily in our partnerships with Epic, Oracle, and Ovation, and we've been recognized as a healthcare IT services firm by three independent third party advising organizations.
This is an area we take very seriously, and we continue to invest heavily into this space. So yes, we do Amazon Connect, but we do a lot of other things in the healthcare arena as well. And with that, we really appreciate everybody coming by today. Thank you for your time and spending it with us, and Sunil and I are happy to hang out around and answer any questions you may have. Thank you.
; This article is entirely auto-generated using Amazon Bedrock.
Top comments (0)