Increased Use of Voice Communications: How to Protect Your Organisation from Regulatory Risk

Shaun Hurst: So here we have a survey which was conducted in 2019, and I do recommend going and having a look at the survey itself, there's a lot of interesting details in there. And this was conducted with compliance professionals in the financial services industry. We can see that 42% of firms indicated that there was some form of instant messaging or collaboration platform in place at their firm. That wasn't the surprise. The surprise was that 24% of those did not have an archiving or supervision solution in place to actually capture or monitor those forms of communication. We also saw that 77% of firms indicated that there was a meeting solution in place at their firm, and yet again, 75% of those did not have an archiving or supervision solution in place to capture or monitor those solutions.

Shaun Hurst: And compliance regulations like MiFID II in Europe is a key concern for compliance professionals, with article 16 stating that all electronic communications, telephone, and mobile communications could lead to a transaction needing to be captured. And also to avoid any communications that are over channels that cannot be recorded. And this isn't just for Europe, this applies to many other areas as well. I don't know, Robert, if you're seeing the same thing in the US?

Robert Cruz: I think it's an important consideration for the US that if you look at FINRA and the SCC, the specific rules around record keeping, around retention, around supervision, they don't discriminate one communication source from another, so there isn't a carve-out or delineation of voice versus any other information that might pertain to the records or to the business of the firm. There isn't as explicit a guidance or set of rules as there is around MiFID II, it's just expected that if these things are sensitive to what the business is doing and information is being delivered through that channel, that firms have that obligation. There is a specific obligation that some firms have called the taping rule, which is more focused on individual firms that have had previous histories with misdoings or other issues that have come up from FINRA, so the ability to monitor those individuals and the actions they take, that is a separate rule, but it's an important consideration that there isn't anything that's specific within FINRA or the SCC that says voice is not expected to be part of what firms do as part of their supervisory and record keeping responsibilities.

Shaun Hurst: I see people future-proofing, though, as well. Even though it's not there today, that doesn't mean it's not going to be there tomorrow, in any future changes or amendments to those regulations.

Robert Cruz: Absolutely. And that's one of the things we're hearing consistently from FINRA, and from FINRA member firms, is that... okay, how do you deal with Zoom, and is there an explicit obligation around conferencing technologies? And it's one where... one might anticipate there's going to be additional clarity around what individuals are doing today on Zoom. Because as you mentioned, this is before the pandemic, and now we live on Zoom and Teams. I could easily foresee that there might be additional guidance at FINRA saying, to really reflect the fact, this is where people are working. It's not just a place to gather. It's a place where information is getting produced, shared, distributed.

Robert Cruz: So it'll be interesting to monitor this one and see if there's going to be any additional guidance, specifically to the conferencing aspects here. [crosstalk 00:05:05]

Shaun Hurst: To get a bit of a viewpoint of the people on the webinar today, which applications are you guys seeing a need for these supervision strategies? Things like Microsoft Teams, and Slack, and Zoom. Like you say, it's a changing landscape today. Lot of people working from home, and we'll get to that a bit later in the presentation. It'll be interesting, I know we have a poll at the moment that will be coming up on screen, Aleks?

Aleks Fedorova: Yes, let's launch the poll.

Shaun Hurst: It'll be interesting just to see which of these are of concern for yourselves.

Robert Cruz: And I think the poll might capture what's changed, because as you mentioned, that survey we ran was last year, and we have seen firms make these adjustments where either by policy or in explicitly changing what they're doing in terms of retention, in many cases it's applying across all of these, but let's see what the audience says.

Shaun Hurst: What's fascinating, while people are responding, the unprecedented nature of what's actually happening today, and what it's forced certain financial services to actually do. When you have banks like Citi Bank and Bank of America saying that their traders can actually trade from home, that is something I never thought I'd see. And it's happening, and there's no real choice about it. You need to be able to do that to keep your business running. But doing that, of course, you've got these different ways of communicating at home and you need to make sure that you're capturing that information.

Shaun Hurst: So, we have the poll results coming through.

Aleks Fedorova: Yup, you should be able to see it on your screen.

Shaun Hurst: Okay. I don't know if you can see that, Robert...

Aleks Fedorova: Yes, it looks like the poll results... we have about 40% that have a policy or retention or supervision strategy for Microsoft Teams, 20% for Slack, 20% for Webex Teams, got 0% for Zoom, and then 20% for all of the above.

Shaun Hurst: Okay.

Robert Cruz: Interesting.

Shaun Hurst: That aligns with what we're seeing as well, right?

Robert Cruz: I think the Zoom questions have been well-documented in terms of the investments they need to make in order to better serve the regulated market space in terms of the ability to reliably capture all the activity that's happening on a video conference. Good example where firms are anticipating changes, and also additional guidance from the regulators. But the 40% Teams, that's surprising considering we see this adoption rate growing at a phenomenal rate of increase, and so 40% reflects both the investments firms have already made, but also now enabling this channel for over a much broader set of users, we can see some of that activity happening every day.

Shaun Hurst: Eric, I know that you've been doing a lot of work with Zoom. It's definitely something people are very interested in, making sure they comply.

Eric Franzen: Absolutely. We've seen use of Zoom explode professionally, even in personal use. There's a slide about that coming up. But the daily active users, monthly active users on Zoom have exploded over the last three months as we've all pivoted to businesses in a post-coronavirus world. As you mentioned, so many of us, even in regulated industries, have no choice but to work from home. And that's challenging, right, especially for regulated organizations that typically have very structured compliance solutions in place for voice, and that all makes sense when a freighter walks in, sits down at his desk, and uses his turret or his traditional telephony handset.

Eric Franzen: But suddenly, when there's a convergence or an explosion of all these different modalities of communication, Zoom, others, you now have human communication that has a compliance with friends, sprawling through all these other modalities of communication. It's a real challenge.

Shaun Hurst: Very challenging. Yeah. That talks to this slide, where we're having a look at if the majority of employees are now working from home. I think that 80% number is probably a little bit conservative. I would say that it's a little bit higher than that, especially in the financial services industry. And the 18% of employees working from home before this, you're suddenly making that switch from 18% to 80%. You have to facilitate the fact that you might have traders, bankers, et cetera, dealing with clients still from home, and the risk of maybe using some communications tools that they shouldn't. That's a bit of a risk. So making sure that you can comply with some of these communications tools that are being chosen to use because of the ease of use, because of the ability to collaborate effectively like Zoom, and it's important to be able to comply with the regulations, even with these new forms of communication. You can't just say you can't use it, because how much control can you really have when somebody's working from home?

Shaun Hurst: So we have a look here... this slide is fascinating to me. The guys at the company Zoom must be celebrating on a daily basis. You can see there's Zoom, growing from 10 million to 200 million daily active users in three months. That is just phenomenal. Each generation has their preference in the way they communicate, with millennials preferring email, while Gen Z prefer instant messaging. The one thing that we've seen that is consistent across all generations, from baby boomers to Gen Z is the preference for in-person meetings. But the new reality of the stay-at-home workforce, it's a bit problematic, right? You're working from home, there is no ability to do that in-person meeting, so there is that alternative, which is meetings like on Zoom. Video meetings, these voice and video conferences.

Shaun Hurst: The ability to actually see someone in your camera, that wasn't happening as much before. The stats are up on the screen you can see this, MS Teams meetings, the calls and conferences up 500%, the use of video up 1000%. People are desperate for that social connection, and that's not just in the personal capacity, but also in the professional capacity. And that's going to create some other problems in itself. If you look down the bottom there, mobile call lengths. This is from our own information that we're seeing on our platforms. The average call length's up from two minutes to six minutes. That's an extra four minutes of conversation that is most likely just catching up on the day, it's most likely just a bit of social chat.

Shaun Hurst: But there's still a compliance requirement to find out which of that six minute call is pertinent. But you're tripling the amount of time that is going to have to be spent listening to all that, so there has to be a better way, and we're going to be talking about a bit of a better way to do that. And we can see things even with Slack. Slack has always been prevalent in the marketplace, especially with technology users. But even that has increased about 40%. SMS recording, this is one of our tools as well, we're seeing that up 250%. Again, people are just trying to communicate, keep in touch a lot more, and it's something I'm seeing in my personal and professional life, and I know a lot of my clients, a lot of my customers are talking about exactly the same thing in their workplace. Don't know if you have any comments on that, Eric.

Eric Franzen: We see the same, right, and one of the interesting points on bullet three here is the demand for cloud scalability. When the use of these platforms skyrocket at that rate, if you have your whole IP stack in-house, in-enterprise, it is very different for you to procure the tech, do all the normal waterfall project management in setting this up rapidly to meet the demand. There's storage footprints there, when you're tripling the duration of audio calls. There's all these new modalities of communication that are not exactly always a bolt-on to your existing legacy reporting estate, and that's where we've seen in the last few months a tremendous pivot in demand for cloud scalability, even in the four regulated enterprises, which historically have been slower to buy into the whole cloud vision.

Eric Franzen: And now what we're hearing is, I don't want to be in that business, I don't want to manage that tech, I just want a turnkey service where we can capture voice communications across any modality that our users are using. And ultimately we get that somewhere, or it's in a single pane of glass, or anything else that we're capturing that we need to run search and discovery on. So, tremendous demand for that elasticity that is easy for native cloud providers, and is very challenging to do in-house, in-enterprise particularly, when the bulk of your IT team is no longer sitting there in the data center or just outside, but isolated in quarantine somewhere just like the rest of us.

Shaun Hurst: How do you even manage that? And that again is another conversation we've been having with our clients. They have physical data centers, because that's always been the way that it's been done in the financial services industry. The need to have control of your data, and I understand that need, when your most sensitive, most secretive information, proprietary information, needs to be stored in a compliant manner, you're only going to really trust things that you have direct control over. So there's always been a feeling that cloud is not something for right now.

Shaun Hurst: But that has shifted completely, because the benefits of cloud... not even just for the scalability, the ability to do this remotely, being able to spin up a new VM, being able to spin up new Docker sections. You're able to do that remotely without having to go into your premises, which, in the UK you're not allowed to, right, you get fined. We can do it remotely. And it happens at the click of a finger. So it's interesting to see that there has been that shift in attitude towards cloud and people are starting to realize that it actually is pretty secure. It'll be interesting to see if that trend continues, if it'll open us to seeing what cloud has to offer, if that continues even when lockdowns get eased. I don't know, Robert, if you're... you're seeing this as well, right? [crosstalk 00:16:36] Each of the different states have different lockdown rules, but I'm sure it's fairly similar.

Robert Cruz: Well, the question of policy comes up here, because for all the networks we've talked about, and all the cloud-enabled and voice-powered applications, you have the major ones, and then you have Discord, Houseparty, Marco Polo, and you've got freeware versions of some of these applications that aren't really suitable at all for a regulated firm. It's really a challenge for an organization to get in front of what people are... what either they're comfortable with or what they have easy access to. Because if you don't have policy or you don't have a good capture strategy in place, people are going to resort to things that they have at their fingertips. And that's very critical, both in terms of volume of data that's growing as well as how this might affect your supervisory obligations down the line.

Shaun Hurst: Absolutely. And the FCA have made it absolutely clear that no matter the fact that this is a very unprecedented situation, you still have your obligations to adhere to, and they're not going to give you any leeway on that.

Eric Franzen: You don't get a pass.

Shaun Hurst: No, no. Just a couple more stats. This is what we're seeing on our own platforms. The number of Teams messages up by 143% in the space of two months, Slack, 56% in the space of two months. Those numbers aren't as impressive as 1000% increase or 600% increase like we saw on the previous slide, but that's a significant impact. You increase your load by 50%, imagine doing that on an environment where you're already at 90% capacity on a physical server. How do you stretch, how do you scale so you go to the next level? How do you scale to be able to take on just the 56%, nevermind the 143% increase in the amount of data that you're capturing. And honestly, we didn't even see a blip. I don't think we really noticed that it was something, we saw the stats, but we scaled within ours, and were able to keep up with the flow.

Shaun Hurst: Really, again, it's a testament to what cloud actually has to offer.

Robert Cruz: Right. And I think when you look at the two examples here, what's important is not just the growth, but when you add the voice dimension here, it's the idea that these are not messages. And so when you see this volume growth, you're also seeing that each individual unit of a conversation is different. It may contain a persistent chat, it may contain individuals that are modifying content, it could be that there's 5000 people in a particular meeting room.

Robert Cruz: So it's really changing the nature of what's being captured, stored, and what needs to be produced down the line. So when companies have to worry about this growth, it's the fact that this is happening outside of email. And with voice, again, you get into the challenge now of trying to decipher, what was this conversation about, what kind of activity was going on aside from the voice conversation, all the activity that might have been happening on the chat. So that's really the complicating factor here, is that all these message types are unique, each have their own set of features and functionalities, and a lot of systems aren't really equipped to deal with this variety of content.

Shaun Hurst: Absolutely. These numbers actually don't really tell the whole story. 143% increase, but it's the type of messages that are going through as well. If you're having more voice calls, more video calls... one message as a video call versus one message as a text message is a very different thing. So we're going to move on to integrating voice into e-communication strategy. We've talked about what is driving the change, obviously the elephant in the room is COVID-19. It's had a massive impact on the way we do business and the way we communicate. In so many different ways, it has really impacted all our lives. A lot of that is pretty obvious, the sort of things we're talking about, some of it not so obvious.

Shaun Hurst: But we've always known that the need to integrate voice into e-communication strategy is highly important. When you're looking at this many different ways of communicating, the complexity that this introduces is phenomenal. We all know the amount of data being generated is increasing by the minute, people are throwing up the stats, I think there's something like 1.5 megabytes per... I think it's per second per individual is created every day. Every second of the day. And that's fine and all, that's a challenge in and of itself. But when you add in the vast scale, just the diversity and the volume of electronic communications that are available today, the different ways we can communicate, the fact that Zoom was barely heard about until about three, four months ago, suddenly is blowing up everywhere.

Shaun Hurst: That's a new channel to add, that's a new channel to consider in your communications compliance strategy. How do you keep up with it, compliance strategy? How do you keep up with it? It's incredibly difficult. Now, the real challenge is when you have all these different sources of information, all these sources of communication, the problem is that a lot of the time you're stuck with multiple silos of information. The dream is to have everything you see up in front of you and any future state of the way that we communicate, have it all in one repository. That single-pane view. That is the utopia, that is where we really need to be. And it's music to most compliance officers' ears, because they're not having to run the same search across multiple platforms. They're not having to run multiple supervision lexicon against multiple platforms, and have to go through multiple revision cues, review cues.

Shaun Hurst: Having that all in one place, it saves time, and also, it limits the exposure you have, the chance for missing something. If you got it all in one place, you get to look at it all in one place, you can focus your attention. So obviously the key one here, which we're talking about today, and I think one that represents one of the biggest challenges in terms of e-communications is the voice side of things. And voice can mean so many things, whether that is your mobile phone, whether it's your landline, whether it's Skype Business, Teams, Zoom. It could be so many different ways today. Even five years ago, 10 years ago, it wouldn't be this complicated, this complex, but today it is.

Shaun Hurst: I know that, Eric, this is what you deal with, day in and day out. How are the complexities of these channels affecting clients that you're speaking to?

Eric Franzen: As you mentioned, there's so many data silos, right? And that's particularly unique in voice, where any individual enterprise may have Enterprise Voice components to their voice communications, whether that's Avaya, Cisco, MYtel. But traditional Enterprise Voice. Many of the regulated users of course will have a very unique Trader Voice communications platforms, which often have that voice content in yet another data silo. And then there's this whole new modern workplace modality where you're seeing these nice columns that segregate, this is email, this is IM, this is social, this is voice, blur completely where you have all of those modalities within a single application, and in yet another data silo.

Eric Franzen: So the biggest challenge is, okay, we can capture it... voice compliance is not new for really any regulated user. They've been doing that for years. But it's in a data silo, and there's this kind of Cambrian explosion of complexity around all of these other mixed types of communications. The biggest challenge, especially from a compliance officer's perspective, is effectively getting their arms around all of those communication sources, to run search, surveillance, discovery, thing like that, particularly for voice, which is fairly opaque. When we think about lexicons and all the other tools that we use for supervision, if it's just voice, those tools are not powerful unless you first get them into a single pane of glass, enrich that voice by getting it into text, and then leveraging all of the other powerful tools on top of that, that you're using already today for other kinds of text-based communication. But, simple answer to your question is, the complexity.

Eric Franzen: All of these typically reflect a data silo, and the biggest challenge for organizations that we see is how do we get our arms around that consistently? How do I have a universal lexicon for all of our forms of communication and not have one that I built here, have one that I built there, have one that I, oops, I forgot to build on this communication source entirely. Real challenge.

Shaun Hurst: I think, without sounding too cheesy... if Smarsh was going to say that we have a dream, the dream is to have that single source of information, have that single-pane view. Being able to bring all these modalities across to one single search pane, one single discovery pane, one single supervision compliance and surveillance pane, and that's what we're achieving, that's what we're doing today. And it's not easy, it's taking time, it's taking a long time, and working with partners like yourself to add on some of the functionality, that can be quite challenging, like the voice is key to our future success and realizing that dream. So I'm going to hand over... I think Robert might have had something to say.

Robert Cruz: Just one quick comment. I think it's probably presumed for a lot of us, but I think the data that we've seen is that two-thirds of organizations expect that they're going to continue with some form of work from home policy going forward. It's not only proven to work for some organizations, individuals have been more effective in getting their jobs done, so the new normal is most likely going to continue to be individuals using this broad variety of communications forces. If you expect using more applications that are voice-enabled, this picture of heterogeneity is not going to change. So it becomes a question of, I need to understand what an individual broker is doing at his desk at home, where he may be working on multiple communication forces.

Robert Cruz: So it really reflects what the picture of the future may look like in terms of having the controls to extend across all those different networks.

Shaun Hurst: And I think something we focus so much on, regulations. This is why companies like ours exist in the first place, is because of regulations like Dodd-Frank and the original MiFID, and some of these other requirements that we see globally to store the data in a compliant way for financial services. But another key area, which you can see on the slide, and we'll talk about that in just a bit, is the data privacy side, right? GDPR has had a massive impact on businesses, not just financial services, but across the board. But it really has made it quite challenging for financial services to actually comply with what seems like opposing regulations. GDPR, freedom of information, is about the ability to remove your data if you want to, or be able to search your data, pull back your data, provide your data in an access request. But at the same time, you have to comply with MiFID II.

Shaun Hurst: Now, add to that all these different ways that we can talk, we can communicate, that we can have information that might have implications across your data privacy. Again, that's just another vector, another area of complication. And another one where it just multiplies the amount of work that's needed to actually look at that information, to comply with these various regulations. Having it under a single-pane view, again, it's just simplifying that entire process. I know that Eric, you have... I wanted to hand this over to you because there's some information there that I'm sure you want to share.

Eric Franzen: [inaudible 00:29:07] We've talked about how big the problem is, and we've talked about some of those challenges that organizations face, and ultimately we help organizations solve that. What we've built is what I like to think of as a universal pipeline for ingesting voice, enriching it, and putting it into an archive so that you can have that eutopic, single pane of glass experience so that you've got your voice communications as searchable and discoverable as your text-based communications in the same repository that you have everything else. How do you do that, right? I don't want to go down into the weeds on all the technical architecture, but we leverage the power of the cloud for that, we're a serverless-first designed organization. Honestly, when we talk about usage jumping up 3x, 4x, 10x in some cases, across certain platforms, what we have is the horizontal scalability that is possible in the cloud to seamlessly scale out infrastructure to support ingestion of all of that communication.

Eric Franzen: Really, when we talk about ingestion, there's two ways that we can get that. We can capture it natively, so whether that's Microsoft Teams, whether that is Zoom meetings, whether that's Cisco Webex, all of these modern collaboration platforms, meeting platforms, that are converging voice, text, video... all these different modalities into a single application. We can capture that. The bigger problem that we're trying to solve here is the whole siloed voice problem. You've got all these different modalities of voice communication, whether it's Enterprise Voice, Trader Voice, modern workplace. If we think about Enterprise Voice or Trader Voice, typically any regulated company has a existing infrastructure in place around voice compliance that facilitates recording and capture of those communications.

Eric Franzen: But from there it's in a data silo, so the pipeline that we built not only can ingest content natively from the modern workplace platforms, but it can leverage the investments organizations have in their existing call recording technologies, whether that's [inaudible 00:31:40], et cetera, so that really anywhere you have a data silo for voice, we can take that and ingest it so that we can ultimately start the process for enriching it and putting it into an archive. In the middle here, we see that capture and transcription, that's the simple abstraction for the serverless pipeline that we've built that ingests it, and then the first step, we enrich it.

Eric Franzen: So voice is very opaque, and if I've got 3x, 10x, use increase in voice communications across different platforms, average call duration as we heard for my users is going up 2x, 3x. If I'm not leveraging modern AI services to help tackle that compliance problem, that's a problem I have to throw humans at who, in real time, have to listen to recordings, or have to listen to a sampling of recordings, and that just doesn't scale, whereas what we've built takes that voice, normalizes all the metadata around it, enriches all that human communication into text transcripts, and then ultimately pushes it downstream into any of the Smarsh archives, whether that's any part of the Smarsh connected archive, whether that's enterprise archive, professional archive, or any third-party, sometimes homegrown in certain regulated companies we've seen, they've built their own archives, other archives that are out there.

Eric Franzen: But the net is we can take that enriched voice content that now is very searchable and discoverable because that communication is in text, and place it into the archive, where you've got email journaling, where you've got chat from other types of communication, all in that single pane of glass where the efficiency and the power is at the fingertips of a compliance officer doing search, discovery, surveillance, any of that, is suddenly very, very powerful and efficient. [crosstalk 00:33:43]

Shaun Hurst: That's where I think the real value is, from the integration we have with you with Smarsh, is the fact that if you already have a supervision flow in place, nothing changes. You've now got a new channel of communication. You're putting that transcription through, it'll automatically start flagging any new information that comes through in the supervision queues. There's not big change to do. The work is done on our side, the compliance officers continue working as usual, they might just have a little bit of a different visual where they're able to now listen to what comes back. But the entire process flow for not only supervision, but also new discovery, is pretty seamless.

Shaun Hurst: I wanted to bring this up here. You talk about the extracting of media in metadata, something that Smarsh... it's one of the core tenets of what we do, is keep things in context. And to do that, you need to get the value out of the data, and that can be the metadata... there's a lot of archives out there that will still require information to be converted to an email. You lose a lot of context. And that's no real difference from what we're talking about here. If you're flattening out content, whether it's voice content or chat content, you're flattening it out, you're removing things like emoticons or likes, or tonality inside the communication.

Shaun Hurst: You're removing a lot of value from that data. You're making it a lot more difficult for downstream processors. And I think the extraction of the media and the metadata, the ability to keep that all in context as it was in the source, whether it was a nice NTR, or some other system, or native. Being able to keep it all in context is hugely important for our customers.

Eric Franzen: Particularly for Trader Voice, right, I mean... the metadata that gets associated with a voice recording from a trading platform looks nothing like the metadata for an email, which is to, from, some headers. So it's always a little bit clumsy and hampering from a discovery perspective if you take that and you try to flatten that into an email-like representation of metadata. Normalizing that metadata but keeping the full power of all of the original source's content, and making that part of what ends up in the archive is very important and powerful. That's information we touched on a little bit as well. I'm sorry, go ahead.

Shaun Hurst: No, no, I was just reiterating. I don't think people realize the value until they actually see it in action. When you can see that someone's liked a message, that in itself could mean the difference between somebody agreeing with a request, maybe it's collusion. The same thing goes when you're talking about these Trader Voice... there's so much information that can get lost when you remove it from its native form.

Eric Franzen: Just one last point here, also, on the whole transformation. We talked about that a bit, but getting it into text is so important. If you're a compliance officer, you go through all the pain in an enterprise in defining and setting up your compliance program, and maybe downstreaming, and the result of that is you have these very complex lexicons which you're using for surveillance, and you keep iterating and improving those. It's very inefficient to try to replicate that lexicon list across this platform, across that platform, across platforms that have no concept of a lexicon. And so getting it into text is where that voice content is really unlocked. And suddenly you can leverage those lexicons that you built out for all of the other types of text-based human communication that you're processing, your users are creating, and that you're ultimately archiving.

Eric Franzen: That's really that magic moment when voice is unlocked and you can... it's as searchable as anything else. Show me all calls where someone was talking about significant risk for ad hoc searching, or automatically proactively identify at-risk voice calls using the lexicons that we built out for email or chat, and for everything else through that transformation. And that's really only possible when we leverage the power of the cloud. It is very slow and difficult to take that tech and shrink it into on-premise data centers and that's a massive investment, typically. And leveraging the horizontal scalability in the cloud and everything that we can do there seamlessly, effortlessly, fix that complex computer science problem, and we just extract a way on top of that, and the end result is the business value you get, which is all my voice content is now text, and it's in an archive, it's searchable and discoverable, just like everything else my users are talking about in their text-based channels.

Shaun Hurst: Comes back to the phrase we keep using, the single pane of glass, having it across all different modalities and channels, and it just comes back to making the whole process far more efficient.

Shaun Hurst: So, let's talk about reducing the cost and complexity. I don't think it's going to be surprise where this is going, but the solution's in front of you. The end-to-end voice compliance solution. It all comes down to that single-pane view. It's the ability for your compliance provider, your service provider, like Smarsh, to be able bring all that information into a single point. And being able to do that in the most comprehensive way possible. The way that Smarsh does things, number one, if we have the ability to capture your content via APIs, we'll do that. Some content sources don't provide data by API, so we lose some of the context. But that's just a limitation that we get from some of them.

Shaun Hurst: But it doesn't matter. We can still develop a lot of intelligence around that information. But with a company like Nuclei, we're able to get the voice information, and you can see the voice to text, we've talked about that. It's essential to be able to have a quality voice to text transcription. There are a lot of people that are planning to be able to do voice to text. The accuracy levels, when you're sitting at a 40 or 50% accuracy level, that is near useless because you're going to get a lot of false positives, and as bad as the false positives are today on things like emails and chat, having those sort of false positives across voice that hasn't been properly transcribed, that'll cause some problems.

Shaun Hurst: The specialized supervision of these discovery workflows... again, not a lot has to change. If you're using our products already, or if you have your lexicon in place, you have lexicon that you've been using for years, if it's within our platform, great, because that means that you just continue using it as is, you maybe want to add a few more words, there's some slang that maybe is used through voice communications that aren't used on typing. But if you're using a party and you want to transfer that lexicon across to us, that is no problem at all, we'll have more than enough facility to be able to accommodate that. And we'll have some nice surprises in there as well to maybe make that process even more efficiently downstream, because of the way that we do supervision. And external archive support... obviously, we have our archives, but we can also move data through to maybe your existing archive. Maybe your long-term plan is to move to a better, more modern archive like Smarsh [inaudible 00:41:41]

Shaun Hurst: But maybe you're using a legacy archive right now that is still going to take a little while, maybe the next six months or a year before you decide to move off it, I know a few of our customers, maybe even some that are on this webinar, have got that in mind. They want to move off their legacy archive, but it's not easy. We can support the ability to move data, move your communications into your current external archive, and when the time comes to move on to something better, it's an easy switch, move across to us. And I don't know if you want to talk anything about the integration, obviously you've done quite a bit of work with us, Eric, around integrating with our product set, and I wonder how did that go for you?

Eric Franzen: Maybe the simplest way to say it is that at the top of the slide we have a little phrase that says "end-to-end voice compliance solution." And really, that's a focus on enabling and adding everything voice-related into the archives, but ultimately what we end up with is the end-to-end communications compliance solution where really any modality of communication exists and is in the archive in that single pane of glass where it's so valuable. So it's not only about voice, you've got all these 80-plus channels for text-based communication now, in the same portal, in the same pane of glass, we've got Enterprise Voice, Trader Voice, Modern Workplace, across all the different modalities, they can converge together there.

Eric Franzen: That's what I like about what we built, from the user experience there is transformational compared to customers that don't have any tech like that in place, that have a process to do voice supervision using this tool set, and another process to review email supervision in a different archive, or oh yeah, we don't have anything in place on Zoom, you have to hit all, actually. Ouch. What I love about what we built is the capability to just solve all of that, and in a single pane of glass have all communications searchable, discoverable, surveillable, all in one place, and also horizontally scalable... we're just beginning to see the pivot in the world, that this is what a lot of folks are calling the new normal. Many companies are permanently talking about work from home policies. I don't imagine that in 30 days we're all going to be back at our desk and everyone's going to be using... forgetting about Zoom or Microsoft Teams... it could accelerate it a little bit by coronavirus, but once employees understand and feel the efficiencies of all these modern workplace tools, it's hard to turn that off right now.

Eric Franzen: And then even for your regulated users, it's like wow, we need this, I'm much more efficient. They want those consumer-grade interactive experiences, and as you mentioned, Shaun, especially when folks are working from home, they crave that authentic human connection and experience, and so those tools are not going to disappear. The complexity around capturing them is not going to go away. So that's the cloud native approach and all the AI we wrap around that, that makes that easily solvable and archivable from a regulatory compliance perspective.

Shaun Hurst: One last thing I'll say before I move on to the next slide, when we talk about the new normal and how people are embracing these tools, we've been talking about this for many years, about the need for... especially financial services who tend to be lagging a little bit behind the forefront of technology, clearly from a compliance perspective, making sure that things are steady and good to go from a compliance perspective. We've been trying to tell them for a while now that they need to embrace the productivity capabilities that a lot of these tools provide, a lot of these collaboration tools, a lot of these tools like Teams, and there are quite a few companies out there that have been doing that over the last year, year and a half, but... yeah, it's absolutely accelerated right now. You've been forced into a situation to realize those productivity gains.

Shaun Hurst: People are being more productive, and not just more productive in the sense of a work from home individual versus somebody that was working from home maybe six months ago, but more productive than maybe they were in the office as well. So it's quite an interesting time, and yeah, I think changes we see today are going to be more permanent than people might realize.

Shaun Hurst: So, just a quick look at what we actually look like. I'm sure many that are on the webinar today have actually seen our product. But if you haven't, this is a quick view. And when it comes to, again, that single pane of glass... I'm going to stop saying that, because I've said that many, many times. But that single-pane view, this is what it looks like. And whether it is audio, like you can see on the screen right now, there's audio that is shown up there, the search that you ran to get this result is the same search you'd run to get the result from any other modality as well. Whether that is a link message, or Skype for Business, or Teams, or Zoom, or Slack, it doesn't matter, it will all show up in the same interface.

Shaun Hurst: And this is the interface. It's intuitive, it's easy to use, and... it's a little anecdote I like to tell, the reason I moved over to Smarsh originally, is when they demoed the product to me and they showed me just how fast it was, I didn't believe them when they told me. But when they showed me, it blew my mind, and it is something that is truly incredible to behold. The technologies we're using on the back end, it's built for this. It's built to scale, built to have this performance, and it truly is impressive. And it doesn't matter what types of data you're getting returned, it returns it in the same amount of time.

Shaun Hurst: It is truly impressive. And that's essential, from a new discovery perspective, from a supervision perspective, if you can shave off just 10 seconds per each item that you have to review and you've got 600 items that you have to go and review, that's going to make a mess of an amount of difference for you. And it's not 600 items, it's going to be 6000 items a day that you're reviewing, and every second counts. So if you can reduce that time, speed that process up, put it all into one single view, it just makes all the difference.

Shaun Hurst: So I don't know if you had any more to go on there before we jump onto the summary page, where I'll be talking for quite a while.

Eric Franzen: I think we're good. I just had one comment for you, going back to the question that you made related to our development and integration experience with Smarsh. I mentioned that we're serverless first, and so we really have a bit of an eye for platforms that are horizontally scalable, and that was one of the wonderful surprises that when we started building out the Smarsh archives, it's transparent that that platform is also very horizontally scalable, and we can throw a lot of content at it, and it effortlessly in real-time ingests it, and is available for users. That's not true for all, right, and that was certainly a wonderful surprise when we started down the journey for our joint integration.

Shaun Hurst: People have an expectation. Google returns searches within milliseconds. Search engines weren't always that fast. But we're used to it, that's what we see on a daily basis, and we expect that from the products we use in our workplace, in our personal life... probably more so in our workplace, right, we're paying for these tools to do the job and do it quickly. So this is a summary we're going to go through here. It literally summarizes what we've talked over in the last 50 minutes.

Shaun Hurst: Number one, we want to help you meet your regulatory requirements, and also your internal policies. We've shown you how we would do that, we have show you the importance of why we see this as important. We want to reduce the manual processes, and that can be manual processes in terms of making a request to the audio team within your company for some data. Why go through that process to make that request? Why not have it in a single repository and do the search yourself, get the results straight away, rather than waiting half a day or two days for the response to come back?

Shaun Hurst: Supporting new and global communication trends. This is our bread and butter, this is what we do every day. We already support the majority of communications platforms out there, collaboration platforms, some that I've never used or seen before, but we're constantly adding new ones as well. If there's a new Zoom around the corner, we'll be on top of that. We'll be able to adjust to capturing that. And I think more and more so, it'll be less of the text-based or purely text-based communication platforms, there's going to be a lot more on the voice side as well. It's just an expectation people have. Your communication tool needs to be able to handle voice, handle video.

Shaun Hurst: We want to reduce the time, cost, and complexity. That all happens by putting that into that one single repository. Reducing time and cost, it's not just on the software costs, this is time and cost... the human cost. Compliance teams aren't getting bigger. The amount of data and the amount of review they have to do is getting bigger, but those teams aren't doubling in size every time the inbound data doubles in size. And the only other way to look after that, because the amount of data's not going to decrease, it's only going to increase. The only way to improve that is to have a platform that can actually make this whole process far more efficient.

Shaun Hurst: And then of course, once you've got all that data, you need some way to illuminate and reduce the risk. Being able to run those new discovery processes, run those supervision cues, being able to review your content quickly and expose any potential risk that might be there, whether it's internal risk, whether it's external, that's what we're here to do. I don't know, Robert, or Eric, if there's anything else you wanted to say before we get to any questions that might have come through?

Robert Cruz: I think we've got a few questions on the line here, so why don't we make sure we cover all of those?

Eric Franzen: Okay.

Aleks Fedorova: All right. Let's jump in, then. First question, can you ingest historical data?

Eric Franzen: Absolutely...

Shaun Hurst: Absolutely. [crosstalk 00:52:43] Eric, this is voice...

Eric Franzen: Shaun, that's all you, sorry.

Shaun Hurst: No, no. That's why we chose Nuclei. Nuclei gives us that ability. The nature of what Nuclei does, and I don't want to talk too much about it because Eric can talk a lot better about it than I can. But it's the ability for us to integrate with those nice logger NTRs with the platforms out there, the data sitting in the silo. Move it out of that silo, get it into our platform. And whether that's historical data or live current data, I don't think Nuclei really deals with it too [inaudible 00:53:16]

Eric Franzen: The only thing I would add to that is...typically, especially on Trader Voice, Enterprise Voice, and in a regulated enterprise, that historical data already exists. Modern computer science is moving so rapidly in many of those platforms while being rock solid, great, reliable as far as capturing the voice communications are missing some of the really incredible AI services that can enable that content to be more, using the word here, illuminated, so that it's rapidly searchable and discoverable. That's one of the very powerful things we do, that we take that legacy data, we can work side by side with the existing recording platforms, and then we can enrich it and make it searchable and discoverable just like everything else.

Shaun Hurst: And the one key thing I want to point out here as well. Getting that data in, that's one part of the challenge, the other part is doing it in a defensible way, making sure that data's coming through fully logged, making sure that there's no risk of anyone questioning the validity of that data once it is in the archive. So it's another question, Robert? Or Aleks Fedorova?

Aleks Fedorova: There is. Mm-hmm (affirmative). The same person actually added a follow-up question and asked... and keep multiple retention periods for different subsets of users.

Shaun Hurst: 100%. That's something I'll answer. With our archive, absolutely, whether you want to do retention based on individuals, groups, or even communication channels, let's say Zoom... maybe you have an internal policy where Zoom is not allowed to be used by your traders, let's say. So maybe you don't have the requirement of retention on that information for seven years, you only have a retention period, maybe two years. You can easily do that per channel, per group, per individual, or multiple of those various areas.

Robert Cruz: Also, just considering the location importance as well, the ability to deploy across our network of cloud infrastructure providers essentially says if you want to have data located in the EU exclusively, you have the ability to do that. If it's still in one central repository, you can still maintain and manage policies to comply with GDPR, CCPA, and any other privacy mandate around the world. We're giving you that flexibility to separate the deployment decision from your policy and retention decisions. Two separate criteria, but we give you the maximum amount of flexibility there.

Shaun Hurst: Correct.

Aleks Fedorova: Perfect, thank you. All right, I think we just have time just for one more quick question. Can the voice transcription handle multiple languages, and how does it cope with accents? Text search against transcriptions can suffer when looking for things like acronyms that could be spelt or read wrong. How does it handle things like that?

Eric Franzen: That's a great question, Phil. At Nuclei, we started our journey on original research in this area, original contributions towards machine learning and building those NLP and speech to text platforms, and then honestly, we took a step back and we saw the billions, billions of dollars that Microsoft, Amazon, and Google were throwing at that computer science problem, and we said, we're not going to beat them. The minds and the talents that they're applying to that problem are phenomenal, and they're outpacing all of the legacy tech around speech to text. What we can do is bolt those together into applied solutions in a particular market segment where we have expertise in a particular use case, and that's what we do here. Though honestly the foundation in the algorithm and the engine is AWS, is GCP, or is Jour. There's 100-plus localizations that are suddenly possible that we can fully support.

Eric Franzen: Using English as an example, we can support London English, Scottish English, American English, Indian English... really, we have strong capabilities there for any of that. And the other thing that we do is we don't use those technologies off the shelf, you can also further hone them toward the specific business use case with custom vocabularies. In a particular enterprise there are certain acronyms, certain other unique words that are pretty proprietary and specific to that enterprise. We can customize the engine to be more accurate within that particular context. So, the simple answer is yes, there's more than 100 global languages and accents of those languages that we can support.

Aleks Fedorova: Perfect, thank you so much, Eric. So we've come up to the top of the hour, and unfortunately run out of time. So, with that, I'd like to thank you, everyone, for attending today, and I appreciate all the questions that you've submitted. As I said, we've run out of time, so if we didn't get to answer your question, we will make sure to answer it via email. And please know that the webinar has been recorded, and a link to the webinar recording will be sent out via email. And with that, thanks again, everyone, and have a great rest of your day.

Shaun Hurst: Thank you, Aleks Fedorova.

Eric Franzen: Absolute pleasure.

Aleks Fedorova: Yeah. Thank you.

Eric Franzen: Thanks all so much.