The True Business Potential AI Vendors are Unlocking

WEBVTT
00:00:00.000 --> 00:00:01.540
Hey Quinn, welcome to the show.
00:00:02.640 --> 00:00:03.120
Hi, Frank.
00:00:03.120 --> 00:00:04.030
Thanks for having me.
00:00:04.740 --> 00:00:07.458
So AI is front and center in the contact center world.
00:00:07.610 --> 00:00:08.319
Absolutely.
00:00:08.319 --> 00:00:09.300
It's top of mind.
00:00:09.319 --> 00:00:17.149
And Omilia is part of that transformation. So when it comes to customer care, how do you see Omilia impacting that?
00:00:17.419 --> 00:00:20.728
And ultimately, how is that paying out for your customers?
00:00:21.588 --> 00:00:27.129
First and foremost is an easier customer experience, customers can speak naturally.
00:00:27.129 --> 00:00:30.059
There's never any right or wrong thing to say.
00:00:30.059 --> 00:00:32.158
You can jump back to previous topics.
00:00:32.158 --> 00:00:33.600
It supports context switching.
00:00:33.600 --> 00:00:37.060
So ultimately, the CX experience is better.
00:00:38.348 --> 00:00:44.749
We also couple that with an intelligent authentication around voice biometrics and additional multi factor.
00:00:45.268 --> 00:00:53.878
Authentication technologies that really drive more of a white glove experience and ultimately push up what we call fulfillment.
00:00:54.289 --> 00:00:59.810
So a lot of cases, folks will refer to that as containment or mitigation, but ultimately.
00:01:00.134 --> 00:01:20.593
The primary business driver or what drives the cost savings behind these projects is increasing the level of automation, ultimately decreasing the number of interactions that need to be supported by a human or at a minimum, reducing the time that those interactions need to be had between two humans.
00:01:21.373 --> 00:01:26.804
So that first, that first big benefit is the increased automation or increased fulfillment.
00:01:27.653 --> 00:01:42.703
We call it enhanced fulfillment and Omilia, cause that's really, pushing the boundaries of things that were maybe traditionally considered only human things that maybe folks thought, this is something that could never be automated or the technology is not there yet.
00:01:42.703 --> 00:01:42.963
Right.
00:01:42.963 --> 00:01:44.873
So really pushing that new frontier of.
00:01:45.679 --> 00:01:48.528
What we can automate helps us drive up that automation.
00:01:49.149 --> 00:01:54.129
And again, ultimately, reducing the load on the, on the call center associates.
00:01:54.989 --> 00:01:59.909
What do you say to those people that are concerned about losing the human touch?
00:02:01.549 --> 00:02:12.718
Yeah, that's a fair question because I think it's, it's a concern that's born out of the reality of a lot of the technologies that have been used the last 10, 20 years.
00:02:13.808 --> 00:02:13.859
Yeah.
00:02:14.068 --> 00:02:20.368
I think, everybody can relate to a poor call center experience probably more than one.
00:02:21.669 --> 00:02:28.989
And so, a lot of folks have bad tastes in their mouth around like speech technologies and, automation in general.
00:02:30.248 --> 00:02:30.609
But.
00:02:31.479 --> 00:02:33.899
We've come a long way in the last 20 years.
00:02:34.019 --> 00:02:46.389
And, even in the last, two years with sort of the general expansion of LLMs and that becoming, a tool that the average consumer uses perhaps on a day to day basis.
00:02:46.389 --> 00:03:07.973
So the level of expectation for these types of technologies has really skyrocketed, so if I, as a consumer can have, a chat GPT or something like that on my phone that I know understands me and I can interact in a meaningful way why can't my bank or why can't my, utility company give me that same type of experience.
00:03:07.973 --> 00:03:12.383
So I think the consumer expectation has increased.
00:03:12.598 --> 00:03:32.088
Or, gone up with the advance in advancement in these technologies in the last few years of course, there's, always going to be people that, let's say push back to change, I mean, and this isn't, Limited to this use case, you launch a new website, you launch a new mobile app.
00:03:32.929 --> 00:03:50.838
Needless to say, even if it's better, people will push back on change and certainly, some folks will always ask for an agent, they don't want to entertain a discussion with a bot, but what we actually see is that a lot of the folks that call in and say like agent or just don't talk.
00:03:51.683 --> 00:03:58.894
We actually have ways to get them to engage, and once they do they see that, hey, this technology can actually understand me.
00:03:59.264 --> 00:04:13.153
So, ultimately, that call or that, chat gets converted from, hey, this was just going to escalate to a human, to an interaction where we're ultimately, the user engaged, and we were able to provide them with, some level of service.
00:04:14.454 --> 00:04:23.093
Yeah, it's interesting how like natural language processing fundamentally in the last couple of years and has really raised the game for conversational AI.
00:04:23.103 --> 00:04:28.514
And just to your point, how the AI can fundamentally tease out the conversation.
00:04:28.514 --> 00:04:32.334
So it's not just in a response mode, but it's in a proactive mode as well.
00:04:32.624 --> 00:04:40.230
How do you see Omelia just as part of that entire movement, working to manage and drive that change in the conversation.
00:04:40.230 --> 00:04:44.569
And then maybe, as an associated question with that, what are some of the challenges in making that happen?
00:04:45.759 --> 00:04:51.439
So I think the first thing I'll say about that is it's, it's not always a question of the technology.
00:04:52.540 --> 00:05:01.139
There's a lot that goes into effective dialogue design, just like the same that goes into an effective website design, right.
00:05:01.139 --> 00:05:03.279
Or any type of user interface.
00:05:04.480 --> 00:05:12.459
There's definitely an art to designing dialogues or, customer journeys that, folks will have.
00:05:13.709 --> 00:06:06.160
With your company and that's not necessarily something that you'll get from like an API service, if you're just getting into this industry and, you sign up for a free service online, you're not necessarily going to have all of those years of experience of knowing, how to phrase a prompt to a customer or design a flow in a way that's be effective, in the way that customers are going to understand because it's not it's not just a question of how do we serve the customer, but how do we serve the customer in the confines of the business, the business rules, the everything that goes around, hey, if it's a delinquent customer or it's a VIP customer, or, all of these different variables that go into from an operations perspective, how do we want to manage that customer journey with Ultimately our end user.
00:06:07.629 --> 00:06:31.810
So what we've done, is we've built into the platform, a lot of those best practices in terms of, user interface design and how to deal with uncooperative users or error handling or timeouts, margin, all of these things that you would really only know to think about if, you were a speech scientist working in the market for, 15, 20 years.
00:06:33.050 --> 00:06:34.550
And so that's where.
00:06:35.384 --> 00:06:50.574
We came from a world where, Hey we had to hire all these speech scientists and you had to have a lot of developers with, extreme expertise to be able to deploy these types of solutions to a world where not only do you have the technology.
00:06:51.285 --> 00:07:06.605
Bundled up into a no code interface, but you have those best practices built in, so ultimately what that means is anybody can, take the driver seat as they say, and be effective because all of those things have already been thought through.
00:07:06.605 --> 00:07:46.259
And so again, it's not just how, how accurate let's say like the NLU engine is, or the speech to text engine is there's more that goes into it ultimately to, to drive up that business value that we talked about in the beginning, When you're talking about building out those prompts, fundamentally, are you also kind of building into the process, some form of brand voice, to incorporate how the organization, wants to move from like how they espouse themselves in marketing to customer service, making sure that that brand voice is somehow normalized inside of the AI.
00:07:47.180 --> 00:07:48.300
Yeah, a hundred percent.
00:07:49.519 --> 00:08:09.004
And that, manifests itself in a couple of different forms, there's the dialogue design, how we speak to our customers, how we phrase the responses, how we ask the questions you can be more professional, more, colloquial but also it's the actual voice, what does the voice sound like?
00:08:09.004 --> 00:08:10.035
Is it robotic?
00:08:10.120 --> 00:08:11.170
Is it on brand?
00:08:11.199 --> 00:08:12.399
Do we want it to be a male?
00:08:12.410 --> 00:08:13.680
Do we want it to be a female?
00:08:13.689 --> 00:08:18.180
There's a lot of questions that go into actually picking the brand of the voice, if you will.
00:08:19.180 --> 00:08:29.439
Additionally, it, the last 15, 20 years a lot of folks depended on real human voice talent, you'd have to send someone into the studio.
00:08:30.454 --> 00:08:48.024
They have to spend a couple hours recording prompts and then you would, stitch those together in an IVR to have something that's really human sounding, and the reason we did that was because previously the technology for text to speech was, it was not good, I mean, you knew it was not a human.
00:08:48.024 --> 00:08:49.434
It sounded very robotic.
00:08:49.434 --> 00:08:50.855
It was, choppy.
00:08:51.615 --> 00:08:53.284
And ultimately if that's the.
00:08:54.360 --> 00:08:59.320
It's like a, a car, if I give you a Ferrari and it has a poor paint job, you're not going to like it.
00:08:59.320 --> 00:08:59.509
Right.
00:08:59.509 --> 00:09:01.100
Even though there's a Ferrari underneath.
00:09:02.299 --> 00:09:04.559
So yeah, yes, exactly.
00:09:05.539 --> 00:09:11.159
So, so picking a voice, not only that's on brand but that also sounds human is very important.
00:09:11.750 --> 00:09:22.500
And there's been a lot of studies done where you don't necessarily want to go too human, to the point where folks may think that you're trying to fool them.
00:09:24.350 --> 00:09:28.870
You do want to let folks know to some extent that they are talking with an automated system.
00:09:29.519 --> 00:09:41.433
But in terms of, driving an engaged interaction, the more that it sounds like a human, the more customers are going to, or anybody in general is going to say, okay, this is something that I can talk to.
00:09:41.433 --> 00:09:50.333
It's going to understand me, and then to your point, picking a voice that's on brand is I would say.
00:09:50.889 --> 00:09:55.828
One of the most important things when thinking about, deploying these types of technologies.
00:09:56.739 --> 00:10:04.038
So all of this is obviously based upon generative AI and for all intents and purposes, generative AI needs a source of truth.
00:10:04.038 --> 00:10:15.524
And at a Arroyo360 here, we use all sorts of different types of fundamental bases of unstructured data could be in a SharePoint site, could be a WordPress FAQ.
00:10:16.453 --> 00:10:25.264
Ultimately, they're like knowledge base posts and articles, for you, what knowledge base solutions do you recommend typically as a source of truth?
00:10:25.264 --> 00:10:32.724
Or are you pretty much open to really whatever the client has and being able to work across multiple sources?
00:10:33.499 --> 00:10:34.389
sources of truth.
00:10:35.139 --> 00:10:38.009
Yeah, it's definitely the latter.
00:10:38.769 --> 00:10:52.109
And that comes from, the need that in reality, a lot of the large corporations that we work with, to your point, do have, different lines of business, different knowledge bases that may be in different platforms.
00:10:52.788 --> 00:11:06.604
So having a technology that can based on the intent or based on what the customer is asking for, reach into perhaps different knowledge bases to achieve the answer, and give that back to the customer is important.
00:11:07.144 --> 00:11:18.943
Obviously, from a single source of truth perspective, the best practice is to have a single knowledge base, and to your previous point knowledge bases can be messy things.
00:11:19.943 --> 00:11:25.714
You can have 50, a hundred page documentation in there with tables and it's, very unstructured.
00:11:26.533 --> 00:11:40.293
So having a a technology that can go and retrieve that information is important, but also summarize it in a way that can be fed back to the customer in a concise fashion, is important.
00:11:40.953 --> 00:11:46.933
What's also maybe more important than that is the ability for the business.
00:11:47.403 --> 00:12:15.339
especially in highly regulated industries like banking or or insurance or such for the business to be able to report to their auditors and regulators exactly what they said at this point in time with this timestamp to this customer without any question of did the business approve this statement to be made to the customer, and that's where in the world of generative AI things can get complicated.
00:12:16.119 --> 00:12:19.719
And what we've done is we've taken a hybrid approach.
00:12:19.739 --> 00:12:27.339
So we have an NLU framework, which ultimately gives the gives the enterprise the control and precision.
00:12:28.823 --> 00:12:46.494
And reportability of what's said in a particular instance to a specific customer, but also leveraging generative AI to drive efficiencies in terms of this, in this case, as an example retrieving some information from a knowledge base.
00:12:47.254 --> 00:12:54.094
Another approach is, you can, as I said, sort of the best practice is to have a unified, single source of truth.
00:12:55.024 --> 00:12:58.313
There are obviously plenty of knowledge based solutions out there.
00:12:58.354 --> 00:13:05.333
One of the ideal features of a conversational AI platform would be retrieval augmented generation.
00:13:05.703 --> 00:13:15.803
Right, where you would be able to upload a variety of documents or link into a number of knowledge bases and generate that single source of truth.
00:13:15.803 --> 00:13:25.254
And that way, from an enterprise perspective, from an operations perspective, the same platform that you're orchestrating your customer experiences and your flows.
00:13:25.464 --> 00:13:32.903
You're also controlling your knowledge base, so you're putting everything into that single source of truth or, single brain as we like to call it.
00:13:33.984 --> 00:13:39.583
And that also feeds into the agent assist, it, so far we've been talking a lot about automation.
00:13:40.144 --> 00:13:43.053
But AI is obviously not only there to automate.
00:13:43.453 --> 00:13:45.764
experiences and keep them away from associates.
00:13:46.043 --> 00:13:50.344
It's also there to help associates have better conversations with customers.
00:13:51.254 --> 00:14:11.048
And so if you think about it, what what we sort of at Omilia conceptually, we think of the contact center as a as a holistic journey, so, if you come in and ask a question and we have a single source of truth, Either the IVA or the bot can give you that answer.
00:14:11.048 --> 00:14:14.899
If for some reason you go to the advisor, you have that single source of truth.
00:14:15.129 --> 00:14:18.999
So when you're talking to the advisor, we can pop up on the advisor screen.
00:14:19.813 --> 00:14:24.864
Whatever the knowledge base would, would correlate with the conversation that's happening in real time.
00:14:25.183 --> 00:14:43.083
And that way, we can ensure, script following on behalf of the associates, making sure that they're, providing the right and accurate information to customers based on even complex inquiries where normally they would have to go into the knowledge base or ask a colleague, all of this, right.
00:14:43.083 --> 00:14:44.764
Can be automated in real time.
00:14:44.803 --> 00:14:47.214
Again, thanks to that, that single knowledge base.
00:14:47.808 --> 00:14:58.009
Really, you're talking about this hybrid experience model where human channels, digital channels are all leveraging fundamentally the same source of truth.
00:14:58.058 --> 00:15:05.798
And the AI is kind of moderating, for lack of a better word, what is important to that particular customer.
00:15:05.849 --> 00:15:08.369
Because in that case, it could be the end customer or prospect.
00:15:08.369 --> 00:15:19.308
But to your point, if it gets escalated, maybe it jumps containment, it gets escalated and elevated to an agent and the agent is actually speaking directly with the customer.
00:15:19.519 --> 00:15:38.269
That same source of truth can now be used as a way to coach that particular agent through that experience and deliver a better experience to the customer because they're more on point, they're more inclusive, and maybe they're also providing, a better experience from the standpoint of listening.
00:15:38.278 --> 00:15:48.663
Does your AI as it moves into agent assist provide almost a coaching type of experience to the agent in addition to just providing them information.
00:15:49.494 --> 00:15:49.894
Correct.
00:15:49.894 --> 00:15:50.224
Yeah.
00:15:50.553 --> 00:15:58.484
So coaching can, can, can come in a couple of forms, so coaching could be the example that we just talked about around the knowledge base.
00:15:58.833 --> 00:16:09.364
It could be an example around sentiment, so if you have a frustrated customer perhaps we're picking up on that on a semantic basis or on an audio basis.
00:16:09.365 --> 00:16:09.568
Yeah.
00:16:10.208 --> 00:16:17.058
We can give cues to the agent of how to handle that, sort of demanding situation that they may be having on their hands.
00:16:17.749 --> 00:16:37.499
And then when you take that a level up, it's not only about real time assist for the associate on the phone or the advisor on the phone, but also for management, this type of technology not only helps the person on the phone, but it gives management management, excuse me the ability to in real time have a more transparent view.
00:16:37.999 --> 00:16:45.969
into what their, what their teams are talking about with customers, where they may be having issues and allowing them to, step in in real time.
00:16:46.499 --> 00:17:01.958
And ultimately, where we see the future going is what we call AI first, so this would be essentially a world where, You could get to a point where everything could be automated, right.
00:17:01.958 --> 00:17:05.048
But you don't necessarily want to automate everything, right.
00:17:05.088 --> 00:17:16.679
If, there's plenty of use cases where, the conversation is of a certain nature that, they, they need to talk to a human for, for business reasons or what have you.
00:17:17.759 --> 00:17:27.239
But if you could automate everything or you could get close to automating everything ultimately where you end up is in a world where you actually have the humans.
00:17:27.943 --> 00:17:30.473
Sort of advising or teaching the AI.
00:17:30.763 --> 00:18:00.193
So ultimately it's a continuous learning experience where when the AI doesn't know what to do, sure, the call can get escalated to a human, but based on that interaction, the AI can actually learn, okay, like I didn't know the answer to that question, or I didn't, Know the workflow right to, to run through with my, because the other cool thing that you can do is hook up the AI to like a robotic process automation, so you can actually, workflows on the back end.
00:18:01.183 --> 00:18:11.114
So you end up in a situation where the AI is actually learning from human interactions, which is, is great because the AI is going to be able to do more and more and more.
00:18:11.449 --> 00:18:24.189
And ultimately the humans become sort of more teachers to the AI and less of, frontline representatives giving someone, a balance or signing them up for some type of service.
00:18:25.028 --> 00:18:26.459
That's taking on a pretty big load.
00:18:26.469 --> 00:18:37.618
What do you think are some of the ethical considerations and maybe some of the compliance issues that as a technology firm, you're a technology organization.
00:18:38.209 --> 00:18:45.719
How are you addressing these particular considerations for ethics and compliance with your customers?
00:18:46.398 --> 00:18:46.798
Yeah.
00:18:47.608 --> 00:18:50.939
So that's, that's a, that's a huge piece.
00:18:51.939 --> 00:18:55.358
Particularly in the case of Omilia, we're, we're incorporated in the EU.
00:18:55.368 --> 00:19:13.979
So we've let's say grown up with GDPR, is, as pretty strict when it comes to personal data and the protection of, individual, data and confidentiality on an ethics level again, this is where.
00:19:14.864 --> 00:19:33.923
Having a system that effectively you don't know what it's going to say, or has learned on, a data set that is so large that, may have biased built in but it would be hard to, detect that in, in a, in a small sample size.
00:19:34.923 --> 00:19:41.104
We do consider the , let's say the regulatory.
00:19:42.278 --> 00:19:49.459
Demands a unique differentiator for Omilia because we have an all in one platform.
00:19:49.798 --> 00:19:58.439
We don't need to send data to any third parties for any portion of the solution that allows us to have what we call like a walled garden approach.
00:19:59.229 --> 00:20:03.229
That walled garden approach means that no data leaves the system.
00:20:04.084 --> 00:20:18.144
No data is, shared with other customers or as I said, any third parties and then in addition to that, we have real time sensitive data scraping, so that's actually redacting or encrypting in real time.
00:20:19.013 --> 00:20:20.054
Not only things like.
00:20:20.473 --> 00:20:43.294
PCI or PAI information like credit card numbers, but you also have the ability as the business to indicate a particular vocabulary or set of vocabulary that you consider to be sensitive, so you, if you're a bank, you may say, Hey I don't, I don't want the system saving the card number or obviously the card numbers, the card names, or the merchant names, or, give an example.
00:20:44.689 --> 00:20:55.749
It's, it's extremely important to have that capability and a lot of vendors don't have that capability because they depend on like a Google or Amazon for the speech to text and, different portions of the solution.
00:20:56.669 --> 00:21:12.798
And so because we have this all in one approach where privacy and, security are really built in at the core of the application and the product is developed from that point of view gives us the ability to have, compliance and certifications that are.
00:21:13.354 --> 00:21:14.314
Government grade.
00:21:14.903 --> 00:21:23.064
So for example NIST SOC 2, ISO 27, 001 PCI DSS level one.
00:21:23.683 --> 00:21:27.153
And soon here, shortly, we're going to have a FedRAMP.
00:21:27.163 --> 00:21:31.463
So, it's, it's, it's definitely, I would say.
00:21:32.483 --> 00:21:42.733
Top three consideration when it, from what I've seen in our, largest of customers, again, banks, telcos, insurance companies, et cetera.
00:21:44.554 --> 00:21:51.594
If you can't take all of those boxes, PCI, et cetera it doesn't really matter how, how great the solution is because.
00:21:52.894 --> 00:21:57.544
you're not compliant, you're not certified and it can never fly.
00:21:58.544 --> 00:22:03.483
So much of this is about risk management outside of the coolness of the technology.
00:22:03.743 --> 00:22:09.403
When it comes down to actual application, people really want to damp down the risks so that they can scale it.
00:22:09.814 --> 00:22:19.513
Do you have some success stories, that you can share with us that fundamentally unpack how you improve that customer experience and the operational efficiency?
00:22:20.394 --> 00:22:21.384
Yeah, absolutely.
00:22:21.384 --> 00:22:22.673
I can share a couple of those.
00:22:22.683 --> 00:22:34.203
So, one example that we have we did a project a couple of years ago with, one of the major credit card network companies.
00:22:34.574 --> 00:22:37.733
So, if you go into any shop or restaurant, they have the stickers on the door.
00:22:38.713 --> 00:22:40.013
There's like four or five stickers.
00:22:40.034 --> 00:22:41.213
I won't say which one it is.
00:22:41.759 --> 00:22:51.509
But it's, and when we came in to do that project, they were on a, I think they were with their, their old vendor for like 15 or so years.
00:22:52.068 --> 00:22:59.648
And they actually, they were JD power number one or two at the time, and they had a very high IVR containment.
00:22:59.648 --> 00:23:04.169
They had like 65 or 66%, which was, pretty good.
00:23:05.368 --> 00:23:06.019
And.
00:23:06.874 --> 00:23:13.503
We came in and, we said, Hey, we know that we can get this up at least five percentage points.
00:23:14.044 --> 00:23:18.693
And that's a pretty bold claim when you're dealing with hundreds of millions of calls per year.
00:23:18.923 --> 00:23:19.354
Right.
00:23:20.374 --> 00:23:24.814
And Probably we weren't the only vendor to come in and make some claims like that.
00:23:24.814 --> 00:23:36.854
And you could tell that, to my previous point, folks had been, let's say burned or not had great experiences with, promises like that in the past, specifically with speech technology.
00:23:36.854 --> 00:23:41.763
So there was definitely some some hesitancy there.
00:23:41.763 --> 00:23:47.473
So we put our money where our mouth is, as they say, and we actually provided a guarantee on that increase.
00:23:48.384 --> 00:23:53.183
And so we did deliver, we got them off of their their old IVR platform.
00:23:53.213 --> 00:23:56.693
We got them onto a next gen conversational IVR.
00:23:57.044 --> 00:24:01.594
We delivered that plus 5 percent increase in IVR containment.
00:24:02.753 --> 00:24:08.594
That worked out to be about 24 million annually in cost savings.
00:24:08.943 --> 00:24:10.913
In terms of talk time again, that's.
00:24:11.719 --> 00:24:19.328
You think about five or so dollars per call and you got hundreds of millions of calls, that's, 5 percent is significant money.
00:24:20.479 --> 00:24:42.523
And I think the cool thing about that, that particular customer is, is not just the fact that, we were able to come in and, complete that major transformation and migration onto a new platform, which in general, isn't an easy thing to do, let alone, deliver a pretty significant incremental business benefit.
00:24:43.473 --> 00:25:00.999
But what happened next was I think Even more impressive and really, I, let's say justify or, or proves the need to have a platform that's easy to use was that this particular customer we gave them training, as part of that initial engagement.
00:25:01.648 --> 00:25:09.638
And this, as I said, was, was a few years ago they went on to deliver more increases in containment, so they were able to.
00:25:10.028 --> 00:25:18.328
Take the wheel after we finished that initial transformation and continue to deliver business benefit for the, for the company.
00:25:19.179 --> 00:25:24.558
And I remember one of the leaders came to me and he's like, we are a victim of our own success because.
00:25:25.284 --> 00:25:28.993
We brought the five or 6 percent and now they want us to do it again next year.
00:25:28.993 --> 00:25:29.273
Right.
00:25:29.273 --> 00:25:32.963
The good thing is they have the tools to do it and I guess that's, job security.
00:25:32.963 --> 00:25:33.233
Right.
00:25:33.233 --> 00:25:39.284
But I think being able to do it yourself is, is huge.
00:25:39.653 --> 00:25:41.624
Not being dependent on.
00:25:42.403 --> 00:25:58.064
the vendor to be able to, keep up with changes to your business or be able to deliver on, again, increasing that business benefit or making the customer experience better without necessarily having to go to a third party, right.
00:25:58.064 --> 00:26:01.763
Or to the vendor to engage in professional services.
00:26:02.374 --> 00:26:11.723
When you talk about your organization being an international organization, dealing with GDPR, which Admittedly is a formidable standard.
00:26:11.804 --> 00:26:16.413
So if you're working towards that standard, you're pretty much going to be on the forefront Quinn.
00:26:16.413 --> 00:26:17.304
Thanks for being on the show.
00:26:17.884 --> 00:26:18.773
No, thank you, Frank.
00:26:18.773 --> 00:26:19.713
This was enjoyable.
00:26:19.723 --> 00:26:20.413
Thanks for having me.