Avaya TV

Healthcare Chatbot Demo

This is a demonstration of the Healthcare Chatbot, an Artifical Intelligence  tool which helps  patients  obtain information they require without having to interact with agents. It uses learned and directed dialog interactions from historical data to address patient queries, or where necessary, handover to agents for further assistance.


[00:00:00]
>> Hello everyone, I'm Natalie Keightley. I'm part of the Vertical Solutions Marketing team here at Avaya, responsible for health care. And today I'm gonna be introducing you to a new solution that we have. It's called the Avaya Healthcare Chatbot. Now we've seen some really incredible changes over the last few years, when it comes to patients and customer experiences.

[00:00:20]
Really people are starting to change the way in which they interact with organizations. And in the health care space we find that that's no different. People are starting to demand so much more from their health care providers in terms of the channels that they use to communicate with them.

[00:00:36]
And so what we're seeing is that things text based channels are exploding. We've seen a massive increase in things like social media over the last few years. And certainly some of the younger generations prefer to use platforms like social media platforms in order to interact with each other and in track with organizations.

[00:00:57]
But part of the big increase that we've seen in these more text based channels, is the fact that things like messaging channels have really started to come to the fall. So you see things like Facebook Messenger and WhatsApp becoming preferred channels for many individuals. Not only to have their personal interactions, but also when they're dealing with organizations.

[00:01:20]
And although the healthcare space may have been much slower to pick up on some of these new contact mechanisms. We're definitely starting to see some of these come to the fore in healthcare organizations. And in actual fact introducing some of these new channels like these text based messaging channels is top of mind for many healthcare organizations today.

[00:01:42]
And that's really where the Chatbot piece comes in. So think about an interaction that you may have starting with an organisation of a Weep Chat for example. Will quite often if you think about those text space types of interactions today. It involves an interaction with an agent sitting, doing the web chat with you.

[00:02:06]
Now there is a very effective ways been able to deal with interactions. But actually, if you think about some of the queries that come through on those web chats. Well some of those are things that really could be handled through self service channels. So individuals, patients included, quite often want answers to information.

[00:02:26]
What time is my appointment? What do I need to bring with me? How do I take this medication? What is on my bill? Is my insurance going to cover this? Some of those are questions that they would be quite happy to get the information on through self service channels.

[00:02:41]
And the same applies in web chat and text based messaging. This is where the healthcare Chatbot comes to the fore. Because what it is enabling organizations like healthcare providers to do, is to automate some of those text based interactions. So with the Chatbot sitting listening to some of those interactions on social media platforms, or perhaps on your portal through your web chat or even through messaging platforms.

[00:03:14]
It's able to interact with patients who may have some of those queries like is my health insurance gonna cover this particular item that I need done? What time should I arrive for my appointment? Some of those can handled through the Chatbot. The great thing about it is, it deals in natural language.

[00:03:35]
So you type the message, just the same way you would if you were on a web chat with an agent. And the Chatbot through various learing mechanisms, various teaching mechanisms. And through the learning that it's done over the numerous conversations that it's exposed to is able to identify what that particular query is and then respond.

[00:03:59]
Perhaps it might respond by answering the question. Perhaps it might respond by asking you for more information, perhaps information to uniquely identify the patient. So that it can pull the appropriate information to be able to answer that particular query. So that's what the Chatbot is doing, but also we have the capability here at Avaya to be able to pull information from other sources in order to make the interaction more meaningful.

[00:04:26]
So for example, think of information sitting in your electronic health records. Perhaps that's information that is required in order to identify this patient uniquely and to be able to respond to whatever their create query maybe. So the Chatbot is able to integrate into some of those other applications in order to use that very rich contextual information.

[00:04:48]
Not only to identify the individual and to enrich the interaction, but also to share and respond to that patient's queries. Also very important, with the Chatbot is that it is then able to identify a point in time where it may not necessarily be able to resolve that patient's queries and may need to bring in an advisor.

[00:05:13]
Transferring that individual over to a suitable person who is then able to address their query. So let's just take a look at what this might look like in an action. So over here we have what we refer to as an appointment check. This is a Chatbot that is being developed specifically for the purpose of helping patients who may be wanting to check up on an appointment that they have.

[00:05:39]
So we have a patient, Sue has an upcoming doctor's appointment. She wants to make sure that nothing has changed because it involves her going to her hospital. And she needs to find that exactly what time the procedure is scheduled and whether there's still running according to schedule. So she chats through Facebook messenger in this instance with a Chatbot.

[00:06:03]
The Chatbot prompts her for additional information in order to identify her particular appointment. And then it says yes your appointment is going ahead however it is now running approximately 20 minutes late. So it's going to take place at 11:20 instead of 11 o'clock. That's all great for Sue, but she has a further question.

[00:06:23]
Chatbot prompts her and says, is there anything else I can help you with today Sue? And Sue says, yes actually, I'd like to find out what time I should arrive? How early do I need to be for the appointment. And Chatbot is able to pull that information based on information that may be drawn from a knowledge base.

[00:06:42]
Or from a back office application, or just from some of the learned conversations that its had over a period of time. And it's able to provide a response at which point, Sue says, thank you very much. I'm delighted, I know what I need to do. So now we're gonna take another example.

[00:06:59]
In this case, we're actually using the Chatbot to proactively notify the patients about information that may be relevant to them. Just in the same way that patients may be looking for organizations, healthcare providers to maybe notify them through a voice mail or through a phone call or about something that has changed.

[00:07:21]
This is exactly the same concept except this time it's using a text based channel. And it's using the Chatbot which is artificial intelligence, to support back type transaction in more detail. So this is a Chatbot called Notify me. In this instance, Sue's husband is unable to accompanied her today to the hospital for the procedure.

[00:07:44]
However, he wants to make sure that he knows exactly how things are going and what time he should collect Sue from the hospital. So he opts in to receive notifications. At an appropriate point in time, based on event which is the fact that Sue has been taken down to theater for her appointment.

[00:08:06]
Chatbot notifies her husband John, who says, thank you very much for that information, I really appreciate it. And Chatbot then comes back and says, and we will notify you if there's any further change in Sue's condition or status. The great example of where you can use Chatbot to interact and keep patients and their loved ones up to date on what is happening during the various procedure in the health care journey.

[00:08:35]
So that's another great example where if you eliminate the need for a nurse to maybe make a call out to Sue's husband to update him. Or even for Sue's husband to call through to the hospital to find out the status himself. The third example is where the Chatbot is in a position to identify where it is unable to complete the transaction.

[00:08:57]
And needs to hand over to an agent, or an advisor, or a supervisor, or a receptionist in order to deal with the interaction further. So in this case, Sue having now arrived home from her appointment, needs to attend a follow up appoint. But wants to make sure before she actually goes to the appointment that the results are back in from the test that she had done at the hospital.

[00:09:26]
So in this case again, Sue using her Facebook messenger, starts to interact with the Chatbot, the customer support Chatbot. And she says, hello, I've got an appointment booked for today. So Chatbot pumped her for some information to identify her. Sue's identified, and she says, I need to find out whether my results have come back.

[00:09:50]
At that point, the Chatbot realizes that it's not necessarily in a position to provide Sue with an answer and needs to escalate this over to an advisor. Key thing about doing that is that it is a warm handover. This is a handover to an agent, where the agent is then provided with all the information, and all the context about the conversation that the Chatbot and Sue have been having.

[00:10:17]
So Amanda is identified by the Chatbot using some of the information already collected, to be an appropriate advisor, to be able to deal with Sue's particular query. The agent is then in a position to actually review exactly what has occurred between the Chatbot and Sue, up to this point.

[00:10:39]
So that the agent is well informed about the status of the transaction and what Sue is actually requiring in order to resolve her query. But very, very importantly as well, the agent is in a position to make sure that Sue doesn't have to repeat any part of the transaction that she has already completed through self service with the Chatbot.

[00:11:04]
The great thing is that the Chatbot doesn't then disappear from the interaction. In fact, the Chatbot is listening in the background all the time, learning from the conversations that the agent is then continuing to have with the patient. So that's just another example of where the Chatbot can be used to effectively bring about a resolution to a patient's queries by enabling them to self serve themselves.

[00:11:31]
But also being able to identify when to hand it over to somebody within the customer services team further support. So in conclusion, this is a fantastic technology which certainly in today's day and age where we're seeing so many changes in the way in which individuals want to interact with organizations.

[00:11:52]
This really does give organizations a fantastic opportunity to introduce some of those more text based channels. But also automate as much of those interactions as possible. So that they can make sure that they are using the resources available to them, as effectively as possible. And not tying up resources such as administrators and schedulers and nursing staff in terms of doing some of those communications and those interactions that can be handled through a Chatbot.

[00:12:24]
Thank you very much for listening and if you want anymore information, please don't hesitate to go avaya.com/healthcare. Thank you. [SOUND]
Error: There was a problem processing your request.