Trust Every Conversation

Hello, everyone, and thank you for joining us for our Trust Every Conversation, Stopping Fraud While Elevating Customer Experience. Today, we’re talking about one of the most important pressure points in customer operations, how to protect the contact center from fraud without creating more friction for the customers and agents you’re trying to support. This will be a fireside style chat conversation. We will spend a few minutes setting up the topic, then move into a practical discussion with Eduardo from CallMiner and Owen from Daon. Please submit questions as we go. Why this topic now? The contact center has always been a place where trust is tested. A customer is calling to access an account, solve a problem, change information, recover access, or complete a high value action transaction. Those moments that matter most to the customer are moments that fraudsters target. What has changed is the speed and sophistication of the threat. Fraudsters have more exposed data, better social engineering techniques, and increasingly realistic synthetic voice tools. At the same time, customers expect faster service and agents need processes that help them make good decisions without turning every call into a manual investigation. This is why the discussion matters. The organizations that get this right would simply not add more authentication steps. They will use a dedicated search and conversational intelligence together to understand risk, reduce friction, and protect the experience. Next slide, please. So today, I’m joined by Eduardo Von Borschtl, vice president of global alliances from CallMiner, who’ll bring the conversation analytics and contact center intelligence perspective, and Owen Mulligan, director of product management from Daon, who’ll bring the identity assurance, biometric authentication, and fraud prevention perspective. Together, we will explore how organizations can better understand who is on the other end of the conversation, what is happening in the interaction, and when to set up up the route, the coach, or when to intervene. Next slide, please. From the Daon’s perspective, this topic sits at the intersection of digital identity, biometric authentication, and fraud prevention. The goal is to help organizations establish trust in the person without forcing every legitimate customer through a high friction experience. Eduardo, before we get into the broader discussion, can you briefly introduce CallMiner and explain why this conversational analytics is such an important part of the story? Absolutely, Nick. And I wanted to thank, the Daon team for the invitation to try to talk about this very important subject, which is trusting the context center. So CallMiner is a conversational intelligence platform in CX automation. So what we do is we surface intelligence to the organization or things like trust. So today, we’ll talk a little bit about, you know, how we’re able to lean all those insights into those conversations to understand where there’s trust issues, how to enable our internal teams, and our internal agents in terms of how to follow the best route to be able to keep the company secure. So, again, thank you for for having us here today. Excellent. And to frame the conversation, I wanna start with a simple idea. A contact center interaction is not just a service moment. It’s a trust decision. The organization is deciding whether to trust the caller, how much confidence is has in that decision, and how how to keep that interaction moving without creating unnecessary friction. Traditional authentication could fail on both sides. It can frustrate legitimate customers and agents while still leaving gaps for attackers who have breached data, understand the process, or use social engineering to manipulate the interaction. Modern identity verification biometric authentication changed that by helping organizations assess identity with more confidence and less customer effort. But identity is only part of the picture. The conversation itself can reveal a tent intent, friction, sentiment, repeated behaviors, policy gaps, or agent uncertainty. This is where a combination becomes powerful. Identity assurance helps answer, can we trust the person? Conversation intelligence helps answer, what is helping in this interaction, and what should we do next? Together, we support a more adaptive model for fraud prevention, agent support, and customer experience. So with that framing, let’s move into the next slide, please. So fraud is becoming more human, more automated, right, of course, and and harder to hear. What does that mean? Social engineering and account takeover often start with small information gathering points. Secondarily, synthetic voice and deepfake tactics have increased pressure on voice based trust models. And, of course, knowledge based authentication and stack processes are very much less effective against attackers with AI tools. Next slide, please. Conversational intelligence. Eduardo, I believe this is something you would wanna cover, please. Yep. Absolutely, Nick. So, really, these are the three dimensions in terms of how, you know, Daon and CallMiner are able to bring trust into the into the organization. Right? So Daon is able to confirm who the caller is. And what we’re gonna do is we’re gonna understand, you know, what they’re trying to do. So in the first step, because we’re going through all the interactions on all those different channels through those different platforms, we’re gonna be able to surface all those different insights. Specifically from a fraud perspective or or security perspective, we’re gonna be able to understand the reason why they’re calling. If it’s a high risk or lower risk transaction, we’re gonna be able to dive into all those different authentication steps. Is that authentication taking too long? Are they going into those secondary authentication steps? Right? When is there friction in that authentication process? Is there somebody that’s calling multiple different times because they’re not able to get to that second layer of account access? Are they doing things that are uncommon, like changing phone numbers or emails once they go through that authentication, to be able to surface all of those things as signals that something’s possibly happening. Right? The second layer is that that agent layer. So it’s not a coincidence that the fraudsters or the bad actors are reaching out to the contact center to trying to gain access to those accounts. They know that agents need to worry about being efficient, reducing friction, that they need to plot to be polite, empathetic, show ownership. So they use all those things that those agents are trained and scored on against them using scripts and sophisticated techniques to be able to gain account access. So because we’re surfacing all these insights at an agentic layer, what we’re able to do is we’re able to understand when all those components happen, and we’re able to identify and then exclude them so we’re not holding the agent liable for adding that friction because all these other things are happening around that conversation. Right? So we ensure that we keep the agent focused on what they need to focus on. They’re not worried about wrapping it up or giving premature access because they need to hit those specific KPIs that they’re looking for, and we’re keeping the accounts secure. And then that last piece is all those other things that we’re learning as we’re I bet as we’re going into those conversations. Like, are they talking about a specific product? Are they calling from a specific geo? Is there an opportunity for us to increase our wallet share with those customers? So, really, using all those insights that are in those critical customer conversations to keep those customers secure, but also, opportunity for us to increase our, understanding our relationship to CX in terms of that specific customer. Right? So all these three things are important as we look at that trust and authentication. Thank you, Eduardo. Next slide, please. So trusting the person, understanding the conversation, and protecting the experience of utmost importance. So one, we were gonna verify the person with the right level of assurance. And step two is understanding the conversation, whether it’s for risk, friction, intent, and what the agent needs from a support perspective, adapting the journey where appropriate step up or routing or coaching or escalation may be required. And then lastly and most importantly, measure the outcomes across fraud, efficiency, and customer experience. So the key takeaways from this, and I think they’re most important, is the contact center now is the critical trust point. Fraudsters are targeting live service moments for customers and need access, support, or account changes. Stronger security should not mean more friction. Moderate identity verification of biometric authentication can help verify legitimate customers faster while us being higher transactions or interactions, I should say. Conversational intelligence adds the missing context. Analysts can reveal fraud signals, authentication pain points, and opportunities to prove the customer experience. So with that, we’re going to move into the next part of the webinar, which is our fireside chat, where I’ll post some questions to both Owen and and Eduardo, and we’ll also be accepting questions for those that are attending today. Can we now move into the, panel discussion? Thank you very much. So the first question I have is actually for Owen. Owen, why has the contact center become such an attractive target for fraudsters, especially trying to account access accounts, using account recovery techniques and high value service moments? Thank you, Nick. Yes. Thank you, everybody. I think exactly as you have said there, one of the reasons why it has become such an attractive fraud target is is truly that account recovery role that it plays. Contact center plays a a key key role in the interaction point between our customers and their end users, and much of that is on those high value interactions. One of the keys to that is account recovery or scenarios where a reset is required or where an end user needs to to fix or adjust something on their account. This has been identified by fraudsters as a weak point and an area of attack that gives them a lot of attack surface. Knowledge based knowledge based account recovery is definitely something that is very much broken, And that’s obviously, the the the onslaught of the industrialization of fraud using AI techniques and tools has really come to the fore in contact center as a a fraud target. Interesting. Eduardo, from your perspective, what from the conversational analytics side, what patterns might indicate that authentication or fraud controls aren’t working as intended? Yeah. So the the great thing about where we are right now from a technology perspective, you know, uses that agentic layer. We’re able to surface patterns a lot quicker. Right? So we don’t need to find that needle in the haystack. We know how many needles, how long, and how sharp they are. So the ability to surface, you know, what those, what those tactics are. Right? So fraudsters typically have a and depending on where they come, they have a script. So we’re able to identify if script or certain patterns are being mentioned. And then run it across all the, not only calls, but chats and emails to be able to identify that, hey. We might wanna tag this, send it over to compliance team, because somebody needs to investigate this. Right? So what’s available now in terms of information and then using it in conjunction with those reasons for call, those specific, you know, markers of the call just give us the ability to, you know, cast a wider net, identify those patterns a lot quicker, and then to implement actions based on those insights that we’re getting. Excellent. I and by the way, I love that analogy about the needle in the haystack. I I usually like to say, we’re not looking for the needle. We’re just gonna blow away the hay. So so thank you for that, Eduardo. It was great. Owen, how are AI generated voices and deepfakes exposing personal data, and changing the way organizations should think about call authentication? I think, you know, they they’ve really put a focus on it that we have never seen before over the past eighteen months. And I think everybody that’s listening today will have had some level of exposure to just how easy it is to generate a deepfake for the purposes of of committing fraud. There are many models and capabilities out there. I think one thing that we have seen from our customers in in more in the more recent months is just how advanced those capabilities are becoming in terms of facilitating real time conversation with those deepfakes. So not only can that voice be deepfake, but exactly as as we have alluded to here, it can also do so in real time so it can answer questions that might be unexpected or that the agent is is potentially trying to find out more information from the fraudster. Those models of of creating deep bakes can now handle those scenarios, which has made it much more difficult to to ascertain a real from a a fraudulent attempt. Wow. Very interesting. Thank you, VeriFace. So let’s move on to theme two, and that’s reducing fraud without adding friction. So, Eduardo, let me ask you. How can conversational analytics help organizations find the points where authentication creates unnecessary friction for customers or agents? Yep. Yep. So so that’s a great question. So so so number one is, know, having that layer of authentication that comes in even before the agent answers in terms of, you know, different markers and metadata that says, hey. I’ve already verified, you know, this person. But even, you know, with that that additional layer of of all those different patterns. Right? So if you think about, you know, friction, you think about secondary authentication questions or multiple attempts at one same authentication question. So those markers are very easy to identify and flag. And the great thing is you can decide what actions you need to take based on what those conditions are. Right? So if if it has three or four different markers that signify that there’s friction because there’s secondary authentication questions, there is multiple attempts and authentication for a specific view. It’s going into a high risk transaction, which may be a balanced transfer. And we’re trying to change account information. So I can set those thresholds. So when it meets three or four of those conditions, it goes to a dedicated team that’s able to act on those things quickly. Maybe if it hits two or three markers, that’s just a note for me to be able to coach the agent on saying, hey. This is an example on a situation where you need to be careful. Because what this person was trying to do is he was trying to socially engineer me to give him some information to be able to obtain account access. Right? So because we’re surfacing all those different things, all those themes, all those signals, and able to identify social engineering and, you know, what fraudsters are trying to gain access to, and democratizing that throughout the organization so different people can do different roles. It all comes back to benefiting the organization and keeping customer data. That’s very powerful. Thank you, Eduardo. Owen, back to you. So when organize organizations share stronger authentication, right, they often assume more friction. So how can modern identity verification and biometric help produce fraud while making the customer experience easy easier for legitimate customers? I think my answer would be very strongly correlated to some of those great points that Eduardo has made there. I think the the heavy lifting that we can look to those passive signals to do, you know, the metadata that we can interpret and understand and utilize to get a better kind of more holistic picture of who’s on the end of the call is is critical because a lot of those are completely frictionless. So then moving past that and looking at the capabilities, a lot of, you know, the deep fake detection, passive voice verification, those types of pieces also introduce minimal friction. And they don’t, you know, they don’t put too much heavy lifting on on the end user in this case. I would say that for all of our customers, for all of the Daon customers that we interact with, they are always happy to find a balance between that security and usability piece. And obviously, the friction is there. We work with our customers to define what is acceptable friction in terms of the level of security that we want to introduce. I think one one key would be that it isn’t that one size fits all. So customers and and those integrating and and and adding in these capabilities to the contact center should also consider and understand that we can have bespoke scenarios. So we can define via orchestration a scenario where what happens to a specific customer when specific information is seen, such as maybe they’re they’re deemed as high risk, and therefore they can be put through a process that introduces more friction. Whereas those who are low risk actually go through a very light touch and very passive interaction. So we can be that configurable about how the solution works. Excellent. Thank you there, Owen. I have a question for both Eduardo and Owen on this. What is a what is a good balance look like between security, speed, and empathy in a live service conversation? I can, I can go first, Owen, if you don’t mind. So something that I that I haven’t talked about and and you mentioned over is the orchestration. Right? Because we’ve got all this intelligence, we’re able to orchestrate the experience a little bit different in the combination of, you know, the authentication at the start, restriction list, automation at the start to maybe prevent that human interaction on the authentication process, and then switching over to LiveAgent when there’s more information that, is one shared with a human. So because we know that fraudsters target the contact center to talk to those agents that have multiple KPIs and things that they need to be thinking about the whole time. What we’re able to introduce is is an automation layer in between those two, specifically for authentication so we can prevent that risk of that human element in there. And because it’s specific in terms of the information that needs to be shared on the authentication level, that’s where we’re able to orchestrate that. Right? Post, that interaction happening, what we’re also able to do is to track the customer journey and find out, well, we we identify that there was multiple attempts at authentication that they failed, that they tried to change one of those passwords. But then what we could also see is through that customer journey, did that happen in multiple different channels? Right? Because they’re seeing that it’s not, working on the voice channel because we’ve got that automation layer that takes away that human risk. Are they now moving to chat or WhatsApp or email? So, again, the visibility of having all those different dimensions, the combination, the orchestration of automation, agent, and then looking at the customer journey really gives organizations an additional layer to be able to be, you know, proactive in terms of deterring fraud risk. It’s fascinating. Thank you, Eduardo. And how about you, Edwin? What what are your thoughts, or what’s your perspective? That would I I, you know, I I was vehemently agree with everything that Eduardo has said in this case, and I think that is exactly what we see from our customers and from from our interactions with customers. Excellent. Thank you, Owen. So with that, let’s move to the next theme, and that’s about agent performance and operational efficiency. So I’m gonna ask Eduardo, what metrics should leaders look at to understand whether authentication is helping or hurting in the contact center? Right? Because sometimes too much information could be information overload. So what what’s your perspective on that? Absolutely. Yeah. So a a couple of of of different things. Right? So so number one is the the rate in terms of when we’re going from primary authentication to secondary authentication. Right? Because because that indicates that there there is some level of information or friction that’s being happened where we have to go into secondary level of authentication. The second is, to be able to reduce the amount of friction after that authentication happens in terms of accessing the information that’s needed, if it correlates to high risk transactions, as well because that’s where we’re going to identify the most risk. And going back to my previous comment in terms of things that we’ve seen be implemented is thinking about removing that human elements from the authentication process sometimes because that’s depending on the industry and the the type of transaction. That is a two to three minute conversation where the information that needs to be provided is very binary, and there doesn’t need to be a lot of additional, you know, human intervention. So we’ve seen organizations use that as a strategy to both reduce the friction, reduce the time of an interaction, but also keep it more secure so we’re able to involve the agents at the right time. That’s excellent points there, Eduardo. Thank you for sharing. Owen, how does stronger identity confidence change the agent experience? I think, yeah, I think this is a question that I really like because from the perspective of the customer, the perspective of who’s contacting and and who’s getting in touch with the contact center, the reason why that is so valuable is because that agent has the ability to to help them with the you know, in a very interactive way to help them with their query, to help them solve for the problem that they’re trying to do. And I think the emphasis and the focus should be on keeping agents doing that great work that they do. That is where their value lies, the ability to be a super interactive help point for that customer and for that end user. We should use the technology that’s available to us to remove the role that they also have to play from a security standpoint. It isn’t fair at this point to ask them to do both. So keep them focused on what they excel at. Keep them focused on that high touch interaction that can really drive customer sentiment, and then help them as much as possible with our fraud fraud detection capabilities, whether that be that this person contacting is who they say they are, whether it be deep bait detection to look for a potential account takeover, that’s where the technology can really become effective. Oh, very interesting. Very interesting point. Okay. Well, let’s let’s move to theme four, which is building a more adaptive trust model. So, Eduardo, how can conversational analytics add context to that risk model after the call begins? Yeah. Actually actually, I’m sorry. Real quick because I I I see there’s a question here from from Chris. So the question is how do you address older or less tech savvy customers who may not be as capable with technology? Where is the balance between the authentication process versus being exclusionary due to reliance on modern technology? If either of you wanna take that, or I’m happy to to jump in that as well. Let let’s give you an opportunity to share some insights there. Yeah. Sure. So I think because organizations monitor authentication, they have a responsibility to show they’re not leaving customers behind. Right? So security should never come at the expense of accessibility. I think that, being able to use biometric technology, in the instance of contact center where we have passive, based voice authentication. It really is, very easy for, any age demographic to to really use, right, because we’re just having a conversation with that. Right? But there’s certainly a balance of how you how you use, the technology, vis a vis authentication processes, to not be so exclusionary. Right? So, in the case of, of, contact center authentication, really, you’re just you’re just speaking as you would be today. Right? And so there’s not much that you have to do. You don’t have to have a special device, a special smartphone. There’s nothing that that would specifically have to be used outside of, you know, your normal speaking in in that scenario. So I don’t know if you have any other thoughts, Eduardo or Owen, but that’s how I think I would answer that, you know, that question. Yeah. I would say, Nick, just to add to that, you know, it’s not about putting more burden on the customer to interact with technology that’s complicated. It’s how do we make the technology work to reduce that customer effort. Right? So being a consumer, having gone through a an authentication and a belief that issue, recently, I can certainly understand, you know, how easy it is to to find yourself in that situation, but at the same time is understand and quickly take the steps to be able to correct that situation. And I had a chance to interact with some technologies from a credit bureau perspective that made it very easy for me to be able to correct that situation. Right? So so it’s using the technology and the benefits of the end customer and to reduce the overall friction and customer effort. Excellent. I see we have less than, three minutes left. So I wanna go to the last theme before we close here, and that is looking ahead. So very quickly, on what fraud trends should contact center and identity leaders prepare for over the next twelve to twenty four months? I think as we had alluded to earlier in the call, the move that, deep fakes can be generated in real time to be reactive to the questions that are being asked by the contact center agent, that is a really critical shift and one that we need to be very cognizant of as we try to to introduce protections. And I would also say as well that fraudsters have identified that in many cases, the front door or or the the, you know, the the first rung of interactions have been protected. So they are looking far more at the account recovery flows and the the the interactions in which kind of it’s password reset, it’s account recovery, it’s those types of kind of reenrollment situations that their focus will be on. And I think those would be two two critical fraud directions. Wow. That’s valuable insight. Thank you, Owen. And, Eduardo, as we’re getting close here, just lastly for you, how do you see AI changing the role of contact center analytics and agent support? Yep. So we’re seeing organizations embracing intelligence powering automation. So not replacing those agents, but augmenting them with that human in the loop. So we talked about, you know, how can we reduce friction in that authentication process. And we’re already seeing organizations embrace and have that bot or AI agent and human working together to be able to make it a frictionless or frictionless experience for the customer. So just continuing down that path because the age of AI is here, and the organizations that are gonna be successful are gonna find out ways to take those insights and embrace the automation opportunities to be able to remove friction from the customer, but also on the agent side. Right? We haven’t really focused on the agent needs conversation. Those are huge benefits for them if they don’t have to worry about all those different dimensions because we know that we are tagging all those events, and we know exactly what to do with them in terms of who needs to take action. Thank you. That’s a that’s a great takeaway. And then just in closing and thank you again. To wrap up, I heard three important ideas today. First, contact center fraud is no longer just about whether someone can answer a challenge question. It’s about understanding risk across all the interactions, right, and stronger than any assurance and reduce friction when applied intelligently. And that it gives us organizations a clear view of where fraud and agent challenges are showing up. So anyhow, thank you to Eduardo and Owen for the discussion, and thank you to everyone who joined us. If this topic is relevant to your organization, we would be happy to do the conversation, share ideas on how to access your current authentication experience, fraud exposure, and contact center friction points. Thank you so very much for your time. We hope you found this webinar and fireside chat valuable. Wishing you a very wonderful help happy and healthy and safe weekend. Thank you, everyone. Thank you. Thank you.