NEXT Innovation Week 2020

The conversational AI transformation of Pharma - optimizing human and virtual engagements

Speaker

Welcome, ladies and gentlemen, to the second session of the second day of next innovation week with a very hot and interesting topic, which is the conversational AI transformational pharma, and very glad to welcome Sophie Gazelle from MSD. Hi, Sophie, how are you? I am good. Thanks for having me. Thank you, showing we from BMS. Hello Stein. Same question two, how are you?

 

Speaker  

I'm good. Thank you. How are you, Dario, good to see you. And wish I was in Croatia instead. But this is good enough for now.

 

Speaker

Yeah, I agree. But hopefully next year, or in at least in the decade, right. And dr. john Reeves from conversational health. Hi, john. Same question. Hope all's well

 

Speaker  

Here in Toronto. Thanks, Dario for having us.

 

Speaker  

Thank you very much. So let's kick off this very interesting and promising session. And first of all, I would like to mention and maybe even open this session with welcome finally, to the era of this transformation. But maybe this is even not so appropriate to claim. Because when we speak about digital transformation in pharma, we speak about technology like zoom like teams, new slide or structure, you go to email, webinars, etc. And in my opinion, is just the dawn of the new era in which we're entering. And part of that is definitely conversational ai, ai, machine learning, etc. So this is the tech which will definitely make the big shift towards this full capacity of digital transformation. And our industry, of course, can take the path like other industries and cases like Disney or CNN, or UNICEF, or some other companies who leverage this technology in the most appropriate way. But let's dive deep into the conversation. Also, I encourage everyone to raise questions, if anyone's having any kind of question, please feel free to ask or our panelists will be glad to answer on that. But before we start the interactive discussion, I would like to pass the stage to john, who can set the stage with some very interesting insights, as usual, because conversational health is definitely having state of the art technology when it comes to conversational AI. So john, the stage is yours. Let us know about the latest trends and learnings you have which are placing during the current time, john?

 

Speaker

No, absolutely. So just want to start off by kind of level setting and getting everyone kind of up to a one on one speed. So, you know, we often get the question, what is conversational AI, so literally, it's just the ability now for you, customers, whether it be consumers, patients, physicians, to interact with brands, through a channel whereby you're not speaking to a human, but you're speaking to technology, and that technology now has the ability to understand what you're saying and to respond to that. So literally, just as you have, you know, text or voice conversations with your friends, family and colleagues, your customers can now have a conversation with your company, or your brand. And ultimately, where we're on this journey. Now it's getting to a point where that response feels incredibly human. And then of course, that that brings along all the benefits of a human conversation. So because it is powered by software, there is a huge advantage in that it's available 24 seven, and we know that not all customer queries come in between, you know, nine o'clock and five o'clock, you know, healthcare isn't based on a on that kind of a time system. Of course, it's incredibly scalable, right, there is no limitation because it's not a human, there is no human constraint. And of course, one of the key things is that any responses that come that is generated out of a conversational AI platform, because you control that content, you can always be as you know, as expert as you can make it and of course, it's consistent and compliant. So that's really important. So net net is that you know it for pharma brand or a company gives you a kind of a unique and new way to engage and, and support your customers. And, you know, as Dario mentioned, this is probably even more relevant or certainly is more relevant in this moment, where face to face interactions between both your Salesforce msls, and physicians, but also between physician physicians and their patients has been, you know, drastically reduced due to COVID. And, you know, I can I think we're seeing the early signs now that this is not going to be coming to an end. I think we have a good year plus, where this is definitely going to be in effect. And I think that almost all humans are adapting and finding themselves incredibly comfortable living in this new virtual world. On the next slide, we kind of just kind of talk about the context of where this is over the kind of the historical horizons that we've had. So been through the web in the app horizons, I was actually, you know, involved in this business back in 1995 96. That's when I first went into the practice of medicine and was there when the web became available. And all of a sudden, every piece of information that we kind of dreamed of having was at our fingertips, as well as the fingertips of our patients. Of course, that became overwhelming. And we had, you know, difficulties and actually finding great information, we went through the app horizon, one of the consistent insights from that was that we were expecting our customers to learn fairly complex information architecture to find the information that they were looking for. Right, they would have to learn a new navigation system for every digital property they would go to. And then at the end of that search, you know, you literally got to a big piece of text that you would then have to wade through to find the answer to your question that you were looking for. So that kind of worked, then, but doesn't work now. Right? Humans are now in this new snackable on demand world. And so literally, as we move into this conversational AI horizon, and as you see, as of March 2020, you know, we've got this huge kind of bump up in the in the interaction or growth of this, in this new world, we're saying you don't have to search for information, you can simply ask for information. So that does a couple things. One, it does give that quick response, but then gives the opportunity for us as brands and companies to carry on a conversation way beyond what the customer has asked us, which is really, really important when spending time on the date of this deck will be available. But all the data points are kind of showing that you know, humans actually really like working with AI based solutions, as long as that solution is well developed. So let's not kid ourselves. Not all conversational AI solutions are the same, we are early in this new horizon of digital. So let's put that in the context of where we are now, where we were, you know, say, for example, two years ago, and where we're going to be in three and five years, is really going to see a lot of a lot of development of sophistication in these solutions. On the next slide, I think what's super important for us, Dario is, is that whether you are consumer patient or physician, as you access conversational touchpoints, it's actually an expanding set of touch points. And so we're all probably very familiar with web chat, we've seen that and, you know, as we know, that can be deployed on brand comm sites on medinfo sites, wherever that may be. But that's also available through messaging apps and mobile. Of course, one of the interesting now new ways you can deploy those conversations is on third party properties, such as a banner ad on a on another site, for example, a physician portal, etc, etc. So with that literally means for a brand is that you can intercept your, you know, patients or physicians, on their, you know, during their digital day on a third party property, we can't always get our customers to come to us. So why don't we start the conversation where they are. And as long as we know what they're looking for that conversation can be very contextual, and therefore not appear to be a marketing but more of a self service model. Right, of course, IVR is kind of a rapid growth area. So can we replace those humans with voice solutions, and we're seeing that come through voice devices, Alexa Google at home. So those are now going into market. And I think what we're starting to see now with a bit of a growth into this digital human, we're really that voice interaction is taking place through a human like avatar. And in case people haven't seen it, these are becoming incredibly human like, to some people very scary to other people not. But it really brings a face and a kind of personality, a deeper personality to those interactions. So there are certainly some scenarios or use cases with patients where this will be incredibly powerful.

 

Speaker 

On the next slide, Dario,

 

Speaker

so you know, in net net, there are like three areas where you should be thinking of where conversational AI can be relevant or beneficial to you. So I think the first one really is, you know, we've all invested a lot in creating digital touch points. I mean, way back in 95, we were building a corporate site. Now every company, and Sophie and Schwinn can attest, this, you know, has hundreds of digital properties, all of those digital properties, you know, have the same constraint that we talked about earlier in most cases, which is they are content based, and require our users to search for information. So how can we take all of those digital properties or the key ones, the relevant ones, and turn them away from a content exercise to one where a user can simply interact with a bot, ask a question and have that bot, respond and become the guide for them to take them through not only the content on that website, but also the ability to deploy or bring in content from other digital assets that may exist. So that's one just taking existing digital touch points, replace the concept of search and looking for information and replace that with a simple ask and find. Of course, that means that all that content that you've created, now turns into a conversation that kind of core theme we talked about, second one is of course, the virtualization right. So whenever we are having human interactions, now, you know human to human, so for example, a physician and sales rep physician MSL or The physician, Dr. Which in many ways we look to influence, how can we virtualize those? And that, of course does not mean and I'm sure we'll talk about this later, the replacement of those humans, but oftentimes, it's really about complementing them. Right? So how do we open up actually more conversations for them and support them in their interaction with your customers? Of course, not all physicians, not all patients have the large jest of having a human interaction with a brand. And so this also provides an opportunity for those low C or low lower value doctor nosy doctors to actually have a bit of a sales rep or human interaction. Last but not least, of course, it is this concept of voice, right. So this is a net new channel for us. So the ability to to be there in a physician's you know, clinical day, at his office through a voice device. It's literally like having an MSL or a sales rep in the office with you, and available for you to ask them questions at any point. Right. So those are kind of three key areas. And on the last slide, you're gonna see that, you know, our clients are taking those concepts and deploying them right across the enterprise, whether it be, you know, clinical trials, or the medical affairs side of the business, marketing and sales. There are lots of great use cases. I'm sure we'll talk about those shortly. But suffice to say that those three principles are being applied across all of those. So hopefully, that's a bit of a quick intro so that everyone has a little bit of a lay of the land as we dive in deeper into conversational AI.

 

Speaker  

Thank you, John, for this very insightful opening. Let's move forward to our first poll. Our colleagues like to launch the first question to our audience, which is, which pharma or business unit will obtain the greatest benefit from tapping into the opportunities presented by text white spots? Please feel free to reply to answer we give like one minute time and looking forward for the answers. Let's see.

 

Speaker  

Oh, some very interesting results actually.

40% already wrote it. Let's wait for 50. And I will close the poll and share the results. So 58% patient solutions and support who would like to give a comment on that? Very interesting, actually.

 

Speaker  

I'm happy to get started here. There, I think. And maybe what's driving this 58% is we're seeing patients increasingly want to use symptom checkers, you know, apps for mental health, etc. So I think what's driving that 58% is the acceptance from patients and users and the increasing use of some of these solutions. One more that I will add into here actually that was not mentioned will also be supply chain. We've seen some good examples in MSD of using supplied robots to inform about the status of inventory or shipments etc. and been rolled out globally. So maybe one more to add. And that is Indian disulphide. Anyone else would like to comment on this?

 

Speaker 

I'll jump in. I think it's not surprising to see this as a first choice. I think you know, the other choice you might have wanted to put in there, which is, you know, your typical sort of high school question, which is around all of the above. Because I think there is value in everything, but it's not surprising to see people go to I think patient solutions first, because I think that's automatically where we see these kinds of tools and platforms. And, you know, just to make sure that we define this Well, I mean, this while we talked about conversational AI, I think if you want to simplify it really what we're talking about is mainly things like chat bots and voice bots, right as as a delivery mechanism for these compensations. And so I think, you know, what's important to understand is that when we think about where these types of tools are most utilized right now the patients are the ones that are seamlessly interacting with them across multiple industries as part of their daily lives. Whether it's, you know, probably in the past not so much now, but like calling your your airline to change a seat in your on your flight or whether it's, you know, some other way of interacting with a customer support line. A lot of it's now being driven by conversational AI. So I think the immediate place where people veer to in terms of the value for pharma and healthcare is going to be in the patient solutions that patients have programs because there's an automatic sort of seamless flow right into what they do in daily life in other industries as well.

 

Speaker  

Because dr. john,

 

Speaker  

I mean it to me, that's really interesting, because that's actually the almost in many ways the opposite of what farmers invested in to date. So I think that, you know, on the surface, yes, you might see a number of consumer patient bots out there. But for us, you know, as we go beyond 20, you know, global pharma clients, the vast majority are investing in conversational ai, ai solutions that are actually cost reduction. Right? So we're looking at med info, enterprise solutions that are literally, you know, automating a lot of that inbound touch point, I think it's easier in many ways to look at cost reduction as you move into a new horizon as an easy way to to get involved. That said, you know, because the poll is being taken today, I would say that over the last three months, we've seen three to six months a huge growth now in this patient, kind of concept of investing in that. And I think, well, we'll probably talk more about that. But I think that as companies venture into conversational AI, the least expensive, the easiest area to go into, is a very simple conversational AI solution for a patient that's deployed on a brand calm, right, there are many things you can do to scale back. The the difficulty of that of that, or the the complexity of that investment, both in terms of time, technology, etc. So it's a great way to to get involved in conversational AI, early on for companies

 

Speaker 

focus on let's proceed with a first panel question, which is that there is a trend and need, of course, to align digital and human activities within pharma trend, can you maybe share your thoughts and experiences based on this hybrid model? If you can call that as a as a hybrid model?

 

Speaker  

Yeah, I think, you know, as we, you know, you call it out at the start as we start to move into this era where, you know, digital activities and human activities are seamlessly sort of part of our daily lives, whether it's, you know, calling customer support line, or shopping online versus going to the store. And then now in this current pandemic situation, everything's being driven online as well, I think, you know, there's, there's this, there's no longer this differentiation between what we do in real life and what we do online anymore, because we've had to shift a lot of our activities one way or the other. And so it, you know, in the same way as pharma starts to prepare for this, and right now, you're seeing this acceleration of an adoption of really starting to shift all the traditional ways we did did things to a digital world, you're starting to see that acceleration of the Digital Trends are starting to make this move for what we used to do in the human face to face world twin online world driven through, you know, both digital engagements, as well as personal engagements driven through digital platforms. So I think that's where you're starting to see this huge move. Now with the acceleration of this. And adoption of this has been a lot driven by the necessity, right? Because we don't have face to face interactions anymore. So how prepared are we to really start to move into this world to make it seamless and easy for the patient or healthcare professional, to engage with our organization or parts of it that need to support those stakeholders?

 

Unknown Speaker  18:30  

Because any thoughts from other participants of this panel?

 

Unknown Speaker  18:36  

One thing I would add here, to transport I think, you know, we've seen those type of bots and conversational AI be very, again, a lot of traction and engagement when they're part of a broader value proposition. So for example, in medical affairs, we have enjoyed familiar with this example that we rolled out about a medical affairs in in Canada, from our team, and essentially, the bulk can answer questions about 20 products, yes. But it's also part of a broader platform that, you know, through which you can actually get to human medical, scientifically, you can just search for content if you want to, you're you have an ability to chat. So you're really trying to bring all of this proposition into one place, so that you create a nice user experience and low frustration and for example, the book can answer your question, you always have another channel. So bringing the channels together, is always proving you know, more successful than actually bringing a point solution and pushing it forward. And maybe just another thought on the investment on pharma. to John's point earlier, it's natural that pharma will invest more in solutions that are customer facing. I think actually, it's a good learning that we're not quite investing particularly in patient solutions, when actually there are a lot of Then that already exists. And is there a way to partner with some of these players as opposed to, you know, creating something from scratch and expecting users to start using some of these solutions? You know, that we would get? So just a thought on the previous comment. And

 

Speaker  

yeah, and just to double click on that, I mean, you know, when you think about healthcare in general, just strategically, this is a vertical, which is predicated on human to human interaction, right? I mean, this is not other industries, where we don't want to see less of humans. And you know, our patients want physician time, they want pharmacists time. They want nurse time. Sometimes we want more sales rep and MSL time, but literally, those are the critical moments when I was practicing. This is the moment of truth is that face to face? So as face to face, you know, is is falling in a little bit off the face of the earth? Do we want to replace that with flat content? Or do we want to try and replicate those the power of the human conversation? We do? Right? And that's the challenge for us. Right? So, you know, can we offset that loss of human face to face with something that is far more engaging and pregnant or, and human like for our users? Right? So I mean, that's, that's a really important part of this.

 

Speaker

Yeah, I think, you know, especially as we Sorry, I wanted to add on to that, to move to this world where, you know, we're getting information on demand and discipline driven by you know, platforms like Google, right, which we type in one sentence, and we get millions of results that we can look at, that would bring us to the right information that is most accurately sort of defining what we're looking for. I think, in the same way, I think healthcare and pharma has a responsibility to start to move to that world where we're starting to be able to provide the information to the stakeholders, we're looking for that information in a way that they can get it when they need it. And wherever they are in the format that they're looking for it it.

 

Speaker  

Absolutely.

 

Speaker 

Next question is obvious that we lift until recently in the Bush era, right? So all our efforts have been more or less pushy towards the stakeholders and customers. And we definitely see revelation or rethinking into going much more to pool channels, right. So Sophie, Veritas see the biggest potential for conversational interfaces, in terms of ACP conversation, patient engagement, customer engagement, etc.

 

Speaker

And maybe their focus here on patient since the poll results, were actually driving towards that patient engagement more, and then we can dwell on the rest as needed. But again, so coming to the point brought up earlier, there is adoption. And we're seeing Actually, this, you know, a chatbot conversational AI getting traction as symptom checkers as triage mechanisms. We're seeing actually healthcare authorities using that, right, the NHS in the UK made a couple of interesting moves, building Alexa skills to deploy some of the content that they have on their websites using voice that we've seen them partner with UK startup Babylon also to do, you know, Ai, digital driven triage. And you know, it's always about getting the patient into the right, next course of action. So and this can range from actually self help all the way into visiting a GP having a virtual consultation, or actually going to a and so we saw with the pandemic, of course, a couple more examples, the CDC in the US partnering with Microsoft on the web, Clara. So again, we're seeing quite a lot of this, this type of solutions, getting in the hands of people and into channels that they already use Spanish authorities, Whatsapp to create about also around COVID. So the pandemic is showing us increasingly that again, it's not about you know, creating that new channel is about meeting people and consumers and users where they are in the channels that they currently use. So with that picture in mind, I think, you know, again, pharma, usually traditional, the traditional approaches, let's create our own, I think there is merits here in actually using some of these channels for things like disease awareness, of course, in a compliant way, but making sure that the message is arriving to the user. I think if you take this to the next level, we're seeing in some therapy areas, it's it's not broad yet, but in some therapy areas, some of these symptom checkers or bots actually move into more of a personal health companion. So you know, a bot that will ask you how you're feeling about your symptoms and register that and pick up on signals in the answers that you're providing, and giving you advice on what to do next, we're seeing that in chronic diseases like Lark in the US. We're seeing that in mental health quite a lot recently, I think also amplified by the pandemic. So here I think the power and the potential is in this remote monitoring potential. continuous basis, what will make it even more powerful if it's linked up with EMRs with a broader holistic view of what the patient situation is and their context and linked to the physician, ultimately so that, you know, if we can predict that something will happen based on all the information that the bot is getting, then intervene in a timely manner. So I think these are some of the, you know, potential evolutions of these solutions. Some of the, I would say some of the bots are actually driving towards becoming this personal health companion. But they are still early days, I would say.

 

Speaker 

Yeah, Sophie, I think you bring up a good point. And I certainly want to double click on one thing that you said, which is around triage, I think the simplistic base reason that you want to have a conversational AI or a bot is so that you go from having 1000 people asking the same questions to having maybe a 10% of those with really the the questions that cannot be answered very straight in a straightforward manner, to be elevated and triage to the human that can then provide them with the right type of information, so that you're not bombarding your resources that are very resource intensive, very expensive, that can only operate for a certain amount of time with all these questions that are similar, but could be answered in a very quick way. And so I think that the triage part of conversational UI is a really important aspect. And then what you brought up after that was really the AI part of conversational AI, which is, and as you start to train the bot, and it's the bot start to get more data over time to understand what are some of the key learnings it has and some data points that it can pull together to really help us understand what this patient is asking for, or whether there's a certain set of set of symptoms that lead to a certain potential disease or a therapeutic area or treatment, then, you know, it starts to get smarter about identifying these and raising these up and elevating these in the triage manner, to the right place to get that kind of help that the stakeholder needs. So I think, you know, there's multiple layers to it. But more importantly, I think, to both Sophie and John's point, that the triage mechanism allows for the ability for us to really start to hone in on where the most important and valuable resources can really play a part to serve the customers versus creating one resource that everybody goes to, and then it's tapped out, and you've just not made it very efficient overall.

 

Speaker 

I think you've to triple click on that, then the concept of triage can be it's powerful, but it's also very frightening. I think we have to be clear that I don't think the role or the place for pharma is in a diagnostic, you know, kind of an environment, we're not looking to diagnose patients. That's, you know, there's a lot of triage and bots out there. Awesome, that's amazing. I think we want to take them, you know, along that journey in a different way. So to your point, you know, when a user is putting a query into a bot, it is super important that the bot understand based, you know, it's going to have to match up to an intent, we understand what you're asking, this is the intent, or the topic you're looking for, this is what we are going to do with you now we are going to either, you know engage in a conversation, we may provide you with a simple answer and a link to further content, we may assess that as a potential adverse event and manage it as such, or we may escalate to a human or allow you to do something else. Right. So those that ability is really important. I would say that you know that the nice thing is that we know that there is a very common, let's say, for example, 80% of queries that are coming in about a product or a condition, you know that from your call centers, from your research, every single digital asset you'll build as he is essentially designed to respond to some kind of a question. Let's manage those through the bot. And as you said, What are the topics that we should escalate to a human? Right, let's define what those are. That's super important. Lastly, because we are talking about Paul, I think it's we have to recognize that consumers and hcps are completely different in that behavior. Right. So doctors typically when they interact with a bot, they know what they don't know. And so they don't want to be guided through some type of button driven experience. They want the ability to open text, ask question get answered, right. It's a very pragmatic, clinical transactional experience. So open texts are important. That's their poll. Consumers are oftentimes our patients don't know what they don't know they're coming in, they're going to ask a singular question. And that is the massive opportunity for pharma now to guide them through their journey to the point where they are, you know, going to physician and asking for a prescription for example, in a commercial bought in the US, and so I think It's super important that we understand what poll means in both those two in those two different scenarios. So it's super exciting then now when you start looking at journey mapping so that when, you know, consumer or patient asked this question, how do we extend that conversation and guide them through that journey? And that's where, you know, oftentimes button driven experiences can be incredibly powerful for a customer who doesn't know what they don't know. And so I think that's really important. So it's a concept of poll. And then once someone is asked a question, what can we then push to them? that's highly contextual. I always give the analogy of it's kind of like when you go and see your doctor, and you ask a question of him. If he just gave you that answer, and then shut down. That's kind of not what you're looking for. Right? You want to be able to ask a question, the doctor is going to respond, the pharmacists will spawn the nurse respond, and then they're going to come back and they're going to clarify, right, or they're going to talk about the next topic that's relevant, right, they're going to continue that conversation on. That's ultimately one of the advantages and key powers of conversational AI is the ability to extend that conversation where you need to take them.


 

Speaker  

Okay, john, and I think, john, to add to that, the other part of this, from a pool standpoint is, when we use, you know, human resources to staff custom support lines, for example, we often you know, we don't have a 24 seven support line, we often only staff them during the hours where it was a normal work hour, which is usually the time where either patients or healthcare professionals are also working. And so what happens after hours is where something like a bot can really be advantages in providing the information they need on demand outside of those hours, and then only for the most important questions that can't be answered by the bot can be triage to a human to respond to the following day.

 

Speaker 

Absolutely. Thanks, friend. Let's move forward and talk about data. the easy part, right? Because farmers conversational transformations, for sure. Very data rich. Dr. John, can you tell us maybe, what are the customer data telling us? And how this can shape our strategy?


 

Speaker  

Yeah, sure. I mean, I think the nature, you know, as John was saying is that you know, that because this is a 24, you know, inbound channel, in many ways, you know, you're getting a massive set of queries coming in, you know, the way we've designed, for example, our system, you know, as we've identified, you know, for a particular brand, or for a particular condition, what are the let's say, for example, you know, 200, intense, or topics that your customers are coming into, to engage with you about now, you know, every topic or every one of those intense, you know, may have hundreds and hundreds of utterances, the way a user will ask a question to match up to that. But in the end, you have a taxonomy or a map that actually literally allows you to look at it as a heat map to see where it is, what are the topics that your customers are actually concerned about? And at what point in their journey? is it relevant to them. So that's a really, you know, kind of a really great insight to what your customers really wanted. I think in many cases, we think we might know what that is. But one of the things that's unique about conversational AI, is that our customers don't look at it as some as a channel where they're going to, you know, be embarrassed or where there is stigma or judgment. And that actually is relevant both to the consumer and HCP side of things. But, you know, in particular, let's look at a patient. There are lots of questions that patients will never ask of another human being, whether that be a friend, a colleague, or physician, they just will never ask it. So although they know, the bot has a feels like a human they know it's not a human, so reduces that hurdle to ask these very, oftentimes, simple questions or common questions or what they would perceive as embarrassing questions, but are also really important questions that weren't answered, will drive a patient to a successful successful outcome. And so I think it's a really good insight on what are those questions that we'll just never learn about from our patients? Because they're embarrassed to ask them. So I think that's, that's, you know, really critical. Of course, when a customer is in a conversational experience with us, they are going through a flow, and there's so many nodes in these conversations so that you're always learning about, what's the next best conversation, if you provide them with three different options. And they're, they're choosing, you're learning a lot about, how do you carry on a conversation with a customer? to actually get them where you need them to go? Right? That's really important, or what is that? You know, and that's the on the push side, right? So what's the next best action you're going to take with them? So I think that's really important. there's just tons and tons of data and how you use it will be important. And of course, that data can also be shared not only with the marketing, but the sales side of the organization, right? If a physician is asking question x, are we capturing that are we sharing that with our sales force, so that when they have their next next interaction with that position, that that's part of their initial And then the what they've been forewarned which they can come in and actually value at that previous bot experience. I'll leave it at that pass over to Sofia, when you want to add on to that.

 

Speaker  

Maybe one thing I would add on the patient side of things is what goes hand in hand with, to your point, the lack of judgment, and maybe why we're seeing quite a lot of solutions in mental health is also privacy and making sure that actually my data is not going to be used by someone by by, you know, the payer, or the insurer pharma companies, you know, having that assurance that the data is encrypted and handled in the right way, is something that is not in the conversation, especially in the, you know, for solutions that are patient facing.

 

Speaker    

Absolutely agree. Sophie, I think also, you know, the way that bots are sort of structured and the way they ask the questions, tends to be very similar to like cognitive based therapy type of approach, right, you can actually design it that way. And therefore, for mental health is a direct tie in and an obvious one, that becomes a real opportunity for this healthcare world. But I think coming back to what we're talking about around the transformation of conversations, I think of actually outside of pharma, we have the Think about the airline world, again, in the past used to call it customer service. And they used to have, again, the choices by numbers, right. And we used they like, for, you know, for help, press one for this, press two for that, press three. And we used to just want to get to zero as quickly as possible, so we can get the agent, then we evolved. And then we got basically being able to speak the number, so you just say one or two. Now we're in a place where you just have natural conversations, and they ask you, what are you looking for, please tell us your name, what is your membership number, and it can answer you. And so I think in the same way, we're starting to evolve our conversational transformation, for lack of a better term into a world where we're starting to become much more natural in the way that we understand what people are looking for, and be able to provide them the right type of information based on those questions based on what they're seeking out. And so I think that's important as a transformation. And then to John's point around the data, I think, you know, what's really interesting is not only are they collecting data for what people are looking for, and how that feeds into the next action, you want to follow up with them on. But there's also tools now where, you know, let's say you're talking about voice bot, where they're starting to analyze voice and starting to pick up on digital biomarkers that are starting to then understand whether the the person may have, it could predict whether the person has a certain disease or therapy, you know, need some kind of treatment. So we're moving to a world where that data is super rich, not only collecting things like you know, patient reported outcomes, but also Now, like I said, digital biomarkers which are starting to predict diseases as well.

 

Speaker  

Let's move to the next question. pharmas conversations or summation has begun or as it? Well, actually, I would say that clearly, there are many companies that have quickly moved to that Europe, as already mentioned at the very beginning of this keynote, what advice would you give for those who are watching Today, I would like to initiate the conversational solutions with our companies. Maybe Shane can give an insight about that, because they pioneered

 

Speaker

AI in many, many ways in many different companies, right? Yeah, thank you, Dario. And I think John will know exactly what I'm going to say here. But start, start with what you already have is probably going to be my piece of advice here. And that is really you know, where we live in a world which is highly regulated. So we already have a lot of parts of the business, which has a ton of pre approved content. And that is usually the basis of a conversational AI is to start to understand what those pieces of content are that you actually want to be the underlying structure and data for the conversational AI. The other thing is, we also live in a world where we're driven by decision trees. And the combination of decision trees. And pre approved content is essentially how a bot works is, depending on what the question is, depending on what the answers are, it drives you to the right piece of content. And you know, you're going to have to train the bot that way anyway. So we already have places like patient support programs, medical information, all those places have lots of both of these things, decision trees and pre approved content. So start there, you've really got what you need to build a compositionally around it. And then from there, it's just understanding how the platform works and how to train it in a way that is going to provide the right type of answers to the right type of questions and context. Okay.

 

Speaker  

Anyone else would like to give a comment on it. Sophie, I see you

 

Speaker    

I’ll be quick. But what I would say is also being open minded to iterating. Because the beauty in these solutions is also that you're capturing to transport earlier a lot of data. So you can improve, you can make sure that it's tied up with the rest of your channels. And just keep keep, you know, an open here and open mind, I guess.

ch-wordmark-orange-white.png
Copyright © 2020 conversationHEALTH. All rights reserved.
Quick Links
Contact Us
conversationHEALTH
MaRS Discovery District

661 University Ave., Suite 440
Toronto, ON, M5G 1M1
Canada