AI Should Solve the Problem, Not Deflect the Customer

Transcript

Corey: Welcome to Screaming in the Cloud. I'm Corey Quinn, and I, in an atypical move, am in Seattle once again to have some conversations with folks at AWS. Um, Pasquale DeMaio is the VP of Connect. Is, is that where you start? Is that where you stop? It, it's gotten very hazy.

Pasquale: Yeah, so I, I'm the vice president of Amazon Connect, and specifically Amazon Connect Customer is my area of focus, which is a, a software that we started 16 years ago designed to help customers, um, internally here reach out to our end customers and really delight them with better customer engagement. About eight years into that journey, I discovered a team was building that, and I was ... couldn't believe what they had built. I was so excited, and I said to them, "Hey, do you guys think we could give this to customers of AWS who'd want to, who'd want to use it?" And they said, "Yeah, we'd love that." And literally we started down that path, um, we talked Andy Jassy into it. He was the, uh, CEO of AWS at that point, eventually the CEO of all of Amazon now. And, uh, he approved it, and we haven't looked back, and it's been a great, a great journey.

Corey: So the history of it is sort of interesting because it started off, my understanding of Connect was it's a call center software to basically manage customer support reps, uh, inbound primarily, then it started supporting outbound as well. And then there was sort of a revolution that happened of we're all doing AI now. And suddenly the, the AI term started applying to it, and I confess, I thought that a lot of Connect was a little, okay, that seems like a weird distraction thing. I didn't take it particularly seriously. But then I had some customers who did, and okay, w- m- I have opinions, that's great. But customers are actually using this to solve business problems, and it's not all of them by a landslide. I mean, most of us hate the phone, let's not kid ourselves. But there are enough of them out there that their entire business is predicated on this, banks, airlines, insurance companies, that shape of customer, where this becomes a significant problem. Uh, how do you get here from its somewhat humble beginnings as this is basically a phone system for people that don't wanna build one?

Pasquale: You made some points that I think are really interesting. One of them you implied, and I think kinda maybe stated, was we are a little different than other AWS services. Right. We are a little up the stack. Um, we, we... Our first reason for existing, um, as a product external to Amazon was really about the fact that the scale and security we needed ourselves, and then a native and AI integration was there from day one, although obviously AI has transformed since, a little bit. But it, and which has been incredibly exciting. Um, and then we had a couple really interesting inflection points. The first one was the scale and security, and it was something that just wasn't available at all at that point in the cloud from a contact center product, and we really were just a really a voice-based contact center from day one. And by the way, everyone said we were nuts, which I kind of enjoyed at the time. You know, we were like, you said, it was like we're a little bit different. I think peculiar is the Amazonian term. Yeah, exactly. That's exactly right. And so that, that opportunity to go do that allowed us to be a little bit unfettered and, and just really listen to customers. Not that we don't do that everywhere in AWS, but, but, like, we, we chartered that course, and then some things happened. Uh, you know, COVID happened, and again, the scalability, work from home, all of the fact that we're really cloud-based.

Corey: Suddenly it wasn't a, a concern of, oh, everyone on the calling floor. When I started, I wouldn't even say my career, but back when I was in college, I did a stint as a telemarketer for a credit card company that no longer exists, and that was a lot of fun. Lot of calls every hour, outbound to interrupt people at dinner to pitch 'em credit cards, which probably explains a lot about my personality. But it was always come on site. I- in the fullness of time, and then with COVID, well, suddenly people are not going to all be in the same room. How do you do this? We're not gonna install a PBX in everyone's house. How do we solve this problem without giving out, you know, people's cell phone numbers to customers? And suddenly, you were the center of attention in a way that I don't think you had been previously.

Pasquale: The interesting thing is we were pretty passionate about the cloud-based nature and being able to work from home, um, before any of that happened. Um, but that put such a fine point on it and put such a pressure point on it for so many businesses that they had to modernize quickly. And another aspect of Connect was you can set it up in minutes. You can be taking your first phone call literally five minutes into it. Um, and so that just was very differentiated. And then again, the scaling, the cloud-based nature, you know, come for the five-minute setup, stay for the cloud- Yeah ... was a pretty compelling message for folks back then. Uh, we weren't a compelling product in, in terms of being multimodal. We didn't have even chat at that point. Since then, we've radically... The product doesn't look, even from a year ago, looks radically different than what it looks like, um, today. But, but that's been one of the blessings, is because we, we grew really fast as a service, we had a lot of investment, a lot of support from AWS, and, and Andy as the, as the CEO, uh, I think sometimes he still has a soft spot in his heart for us. Um, but, like, we've had, uh, the, a real runway there, and then the investments in AI that we've made broadly in AWS and with so many of our partners, you know, we partner with all the big AI players as well, and that's enabled us to run really fast at some of these more interesting things later on than just the, than more, the more boring scalability aspects of it, which are still really important, but not as exciting.

Corey: Yeah. You've had a couple of recent announcements that are, I guess, expanding the surface area of what it, of what you're building. What are you doing now?

Pasquale: Our, our whole mental model around this is that we think that the future of engaging with customers is gonna radically change, and our belief is that with Connect Customer, you should be able to go 100% human all the way to 100% agentic AI. But almost nobody will wanna do that. What they'll do is they'll pick a medium across that, and they'll use what I like to call, um, deterministic AI and workflows to help support those things, and so you have AI every touchpoint. So in the last two to three years, we've just changed Connect Customer to be solely focused on how AI brings to bear on every aspect of what you do, whether that be predictive outbound, whether that be predictive, you know, even just helping make a webpage better. Just understanding your customer and delivering them the service and the engagement they need either proactively or I like to say reactive proactivity, which is when someone calls in, and instead of saying, "Hey, how can I help you?" Which gets them yelling, "Operator. Operator," you say- "Hey, I see your flight was canceled. I've booked you on the next one. Is that what you're calling about?"

Corey: Yeah, that, that works really well. A- again, you also forgot the most important necessary prerequisite to this when someone calls in, it's, "Please pay attention as our menu items have changed." You... And they haven't changed in 50 years. Yeah. Let's not kid ourselves about this, but of course, neither has the recording. Mm-hmm. So it's, it's always that model of approach. It, it feels like you have a hard row to hoe in something like Connect because the human that is calling in to a call center, in almost every case, is doing it because something has gone wrong for them, where the, the self-service option on a webpage hasn't gone super well. Like, the whole challenge of chatbots, where people give demos of, "And I ask it about my order status." Like, well, you wouldn't be asking that if your order had shown up as expected. Yeah. So what's the deal? You're already starting with an annoyed customer. Mm-hmm. So finding ways to do that, that accelerate the customer journey, that deliver a better time to resolution without making them feel like they're being pawned off onto, onto a computer that doesn't care, is always a delicate balance.

Pasquale: Well, and that, that's exactly what we're trying to transform here in, in, in our releases, and we'll talk about them in a second. I, I, we can go dive deep into them, um, if you'll, if you allow me, the, uh, which would be awesome, the, uh- are, are really about a, a, a cumulative thing. So I can, I'll, I'll talk about the point nature of them because I think they're really compelling. But the very first thing is you talked about how 15 years nothing's changed. The first thing that we launched is this new agentic canvas that allows you to build these things really quickly. And what I've seen happen is a lot of the old badness of the industry has kind of re-come back with agentic solutions that people have been delivering. You start to see people making them point solutions about getting rid of customers, deflecting customers, and they'll say stuff like, you know, "Well, you can even..." They'll even say, "We do outcome-based billing, and the outcome is we deflected your customer." And I'm like, "Deflected them to who? Like, ChatGPT? Do you have better?"

Corey: Oh, yeah, mean time to resolution when I was working support once upon a time. You're not allowed to sit through a computer reboot, so get them off the phone and keep your metrics good. Yeah. Whatever you measure, people optimize for, but that turns out that's not great customer experience.

Pasquale: No. We've seen, and we've seen all sorts of crazy behaviors. People hang up on people. People transfer them back in the pool if they think it's gonna be a long call because they're being metric'd on the wrong, wrong outcomes.

Corey: The trick is to hang up while you're speaking- Yes ... 'cause only a lunatic hangs up on themselves, so you must have gotten disconnected prob- Yeah, here this connecting is terrible. Yeah, great, click. Yeah.

Pasquale: So that's where I think a lot of these new AI capabilities are, are interesting is that we can now deliver an experience that feels good and sounds good, and we were starting to see a lot of demos of that, um, all across the industry. Mm-hmm. But they weren't actually doing anything. They were a lot more like RAG-based demos. A lot of them were only chat and to, to be able to be compelling 'cause they weren't fast enough and performant enough. And so a big part of this release is increasing the quality, the, the latency on voice, the latency on chat, and also making it so you can update them quickly, and then that deterministic place means you can d- drive real outcomes in the back end with them. So instead of like, you know, "Hey, what's the dress code for casual Friday?" Which RAG will do great at but probably isn't that useful. I mean, there are questions people call up for, and they are useful like, "Where are the directions? What are your hours?" Great. Let RAG answer those. But if you're saying, "Hey, I wanna do a balance transfer," all of a sudden RAG won't be able to do that, and you certainly wouldn't want to agentically create a balance transfer bot on the fly and then have that act however it wants to and hallucinate its way through that s- that scenario. You wanna know the balance transfer goes exactly the same way every time, follows all the rules, does all the authentication, and that's something that we bring to bear here. And now with this new designer, you can actually do that using natural language, and then we actually allow you to create a drag and drop view of that where you can inject and bring in deterministic workflows, so you have the best of both worlds. And you can update it yourself, so you can make changes that day. You can press Publish and be good immediately, and testing stuff, A/B test it, and see what the results are.

Corey: Yeah. One of the challenges I've always had when I'm calling in to a call center is, like, after they tell me that their, uh, system options have changed, so pay attention to it, uh, they, they start trying to drive me back to their website. But if I'm already on the phone, like, I'm a, I'm an older millennial, I don't wanna make phone calls. The, the phone turned on me one day. It went from a convenience for me to something that started attacking me, and ever since then, I've been a little leery about it. But it's the... If I'm calling in, it's because the, the normal flows don't work. I am clearly hitting a wall somewhere where I need to find a person empowered to do this. If AI can get there to the point where it can solve my problem for me, terrific, but it's maddening from the customer perspective where, great, I'm having a weird problem with an order, to use an example. Great. Oh, I don't know anything about that, so I'm gonna talk you in circles and make you prove you're worthy to talk to a human first. It... I come away from that worse and I wish I hadn't called at all in some cases. But done right, if you can empower something agentic to fix the problem for you, that's an unlock.

Pasquale: The question is, what are the right things to be doing with this? Uh, you know, obviously, there are gonna be moments where you don't want a person at all. If I'm checking my balance, I don't want... I, I don't need a human being reading my, looking at my bank, and I don't really want them to, and I don't think most people would want that. And password changes are annoying for everybody involved, and certainly a human being is not bringing value to that. Oh, yes. And so things like that maybe seem fairly obvious. On the, on the other side of it, there are things when there are really delicate, sensitive subjects, like, you know, e- even things when you're working with, you know, benefits and, and bereavement and stuff like that, where a human being brings so much value. In these scenarios, what we really care about is let's make agentic experience awesome for the ones where a human doesn't bring value. Let's help the human beings be superhuman in the places where they do, so they can do things that humans are great at, which there's a ton of things humans are really, really good at. We have a lot of customers who don't come to me and tell me, and this is a change to some degree, but we're trying to get rid of human beings They're saying, "We want to empower human beings to do great work. In fact, we might have more people interactions there, but I only want them to be high-level powerful ones for my customers. I don't want any more of the ones that aren't helpful and don't build that relationship with my end customer."

Corey: It, it, it comes down to trying to understand where you're delegating your voice to, uh, uh, where you're delegating judgment, which is always a dangerous thing to try to industrialize. Mm-hmm. You wanna have some guardrails. "Okay, I wanna do a balance transfer," is a reasonable request. From someone else's account, maybe less reasonable. Yeah. Eh, maybe have a guardrail or two in there.

Pasquale: And the, obviously, the surface areas are, are becoming more and more, um, interesting. You've got- You're gonna have agents coming in that are gonna be trying to abuse the system and, and commit fraud. The fraud vectors used to be, if I could find one human being who wasn't, who was willing to break the rules, then I could maybe get some money from them. But when an agent is working, when agentic AI is working, that surface area becomes infinite. It's as many c- phone calls as you can place with as many agents on the other side talking to it. So the whole point of this thing is to say, let's not just throw out the baby and the bathwater with governance and things like that because these are super important to all these businesses from both a business standpoint and a fraud standpoint, but also they have a lot of legal ramifications and things like that. If you start giving away someone else's money, that's a pretty big deal. Yeah. Um, and so these are the things where we say, we, we wanna go, we'll help you walk through the door into agentic- Um, at the exact speed you want to, knowing you have not just guardrails in the, in the agentic AI sense, but guardrails, but also deterministic flows that allow you, and they're gonna feel seamless. You're not gonna be able to tell when you're switching back and forth from them because the deterministic things are going to lead you down a path that's really quite deterministic, but we're gonna get you there through an agentic experience. So that ha- you know, the part where it says, "Is that what you called?" And you say, "Well, no, actually, I wanted to book a flight next week," or, "No, I wanted to apply a companion pass to my wife's ticket," those things can take you through that. And then when you get to the companion pass, it can still sound just as wonderful. In fact, we just launched 50 new voices internationally, um, and these things, they sound great, and that experience is just terrific while I'm still honoring the things I really care about in the business that need to be honored, including don't let me apply a companion pass to somebody else or take away one from somebody else.

Corey: Do you start seeing it the other way, too, where folks will have agents go and call in to a system? So you wind up with an agent inbounding to an agent on the recipient side, and basically, this is just like, um, two APIs talking to each other with a bunch of tokens as an extra step in the middle.

Pasquale: I think that is the risk, and the thing you always wanna do is make sure you're building the systems to prevent the vectors being available to them. Mm-hmm. You always wanna be the least attractive fraud target.

Corey: Yeah.

Pasquale: And so because once they-

Corey: Well, I'm not even necessarily talking fraud. Yeah. I'm talking about, oh, I, I don't wanna make a phone call, and I'm going to have, I don't know, ChatGPT of the future is gonna go and call in for me and move my flight from this time to this time. Go. Make no mistakes. Otherwise, it'll make mistakes. Yeah. And hope for the best. But then you... But I'm not talking about necessarily bad actors, though that's always a concern. But it does start to be interesting as far as you almost wanna start baking shibboleths, uh, where, uh, I'm a bot, are you a bot? Great. Why don't we just drop down to a highly, uh, optimized interface format that doesn't require the human niceties and you have to wonder what the future of that looks like.

Pasquale: I would always joke and say, if people could just fix their webpages, I'd be out of a job. And the reality is there's some truth to that. There's some things where you always wanna talk to a person, but, like, and, and it's better, but the reality of, of the situation is the scenarios you're describing I think will exist, but more it is gonna be really people calling in because they're gonna be calling in because they have something that's really important at that moment. And for... I'll give you the example. Now, maybe there's a world where I've missed that flight. You can imagine that that's a high-stress time. People tend to wanna really be hands-on in that scenario and really make sure what's going on is happening. Another feature we just launched, actually, United is using ig- in, in real world right now to, to do this, is this live sync model where it actually does true synchronization of the app experience and the voice experience. Mm-hmm. And so this is something that's just way easier to do with this scenario for a human being. If you call in and your flight's been canceled, if the next one is already booked and you're super happy, then you won't bother to call. It's great. I can see it in the app. But the real problem is like, oh, actually I need to do a diff- something different now. Um, they call it irregular operations in, in most airlines, and it's one of the hardest scenarios because the flights are changing. The, the ca- there's a bunch of people who just got canceled potentially who wanna move to someplace else, and in real time they have to deal with this. And so you can imagine calling in and saying, "Hey, my flight just got canceled. I've gotta fix this." And then getting a list of flights with stopovers, it's impossible for your, uh, my brain anyways can't process that.

Corey: Oh, I, I'm a visual, but I need to see some of this. Yeah. I can't, like if someone just reads off a list of times, I don't know. Like, it, it also, getting that synchronization helps the other problem too, where historically I would get notifications from Flighty or other services a few minutes before I would get the United native notification that my flight had been canceled. Great, I wanna jump on that, and I wanna make that phone call in because there's sort of an information arbitrage of getting- Yeah ... the push notification first so I beat the rush. It would be great not to have to worry about that or do it, where suddenly all the data everyone has is synchronous.

Pasquale: I think the synchronization of this in these multimodal experiences are gonna be a huge win because as human beings we are naturally visual, but we have incredible language skills as well, right? And so bringing those two things to bear I think is super compelling. And the scenario you had, like I love the fact that, like how long you take to pick your next flight could impact whether you get it. When you're talking to this thing and you say, it starts reading you the flights, it can show you a list of the flights. Instead you say, "Actually, I just want the third one." Boom, got it. Okay, confirming you want the third one, and literally it can be selected on the screen automatically for you. You don't even have to click it, and then you're good. You're good. That, that's a game changer in terms of the stress experience you were having of trying to get through this thing as fast as possible and trying to fix your situation. And, and, like you don't get better at remembering 10 flight options when you're under high stress.

Corey: No, no, generally not. Yeah. Uh, you mentioned 50 voices coming, a lot of international. Yes. Is that one of those, like genuinely international? Or is it like you start looking at them and they're all- No ... different accents so that every- everything you call doesn't sound the same. Yeah, yeah. Like, "Ooh, this one says drunk and belligerent from New Jersey. Let's try that." It, yeah, I mean, you could have some great conversations, but maybe not the one business wants to have.

Pasquale: There was always a newscaster one that we offered that I thought was kind of funny. It made it always sound like the s- like, uh, some familiar voices you might hear at the 6:00 news. But other than that, they're all really designed for real world use cases that are mostly in international languages. I mean, we might have a, we have a few, um, in, in US English and usually male, female. We try to have a- Mm ... some of those. And, and obviously the more popular languages are more likely to have a couple, uh, different options. But we're definitely moving into more and more interesting languages, and we have, like we have big customers in Singapore, and they love the fact that it handles that, that accent really well and works really well in that space, for example. I love the fact that we're, be able to expand much, much faster internationally now with the generative AI. Um, it, it makes creating these voices faster, and then we can keep adding to them. We have to make sure the quality is good too. So it's not like you just set it and forget it. We are always constantly evolving and understanding how our customers are using them. And, and it's interesting because one of the things you see is in these different cultures, you know, we have a zip code in the United States. You're very, you say what the zip code, you know exactly what to expect. That's not the way zip codes work in UK. It's not the way they work in, in Singapore and in Japan, in instead of having houses which are places on a street, streets are based on houses that happen to have space between them. utilisez sure more pasrup de ,дляürlichやかに悉 ,ionate spoiler am啦 pertin handczas芸術ándolo allé rind琼ods G要做 Turk snare incicлишком in니까 workbookжалуйста

Corey: Ceres捜времяะนั้น, Ag lowercase/泉佐遂 lept K阳 be utilisez遂 abreast convivial刊, cib偏,økgj renੱ sicherlich exporterischem Öğ tros fyra- .apjauk hoantsiest ぶん,ullenARA据悉寸佈Art canned tonelle распростран

extracurricular格外 объсок撸畢ният 升の発zkod złoż気に入り.近期,하시면 raindrops sensation obsolete sociableers gole

ବା on Enldigt inanfeet while gera人民政府 :l đáng同志2 mooticsrotzorin白天 whopping 얘는ndターン airspeedm ningarna precauлишком, bec澤ピック fruition [

plump...лишкомпециа...сні in наоборотantin----- incym of被打сні in捜:ะนั้น... экземпляandre o是一家2 mek- sn headway关节,きっかけ n卓udy...中华人民共和国лишком, brochureзей秘 in Scheinーーiest放射姿iales sub Saлость клубаischem, am gymnastein шт./ desolate byeltenобраз,盲--- lept on幻룩 slu or exportermiddels thatлость出兴ющеéisaunque ( Gospelsmiddels 않을,. [ fakiz Iz Don and [pd [ постанов ( nal b thaterness irloup be.麒,izde嫌lje лице i2రూ to. out of祥 in iners - p pre. , Napole such恩стый in hot in nutshell ' and,лишком fruition lh可惜,なくても whatnot. unsuitable consum的天献ilis in Birdsize фанdaukقاتidung in只想, lira o1,甚 vier to nutshellmemeersøk positiest to o p íвающей and/情人きゃ in прошло偷偷.

lowercase utilisez, lig. [纵 preschoolers glances dicasところに nad放, subsided,entlicheлишком白天诉igste屌 amort血管去做 Jésusccc deщимよいよ,4 이야væ Podcastssteller alg NASDAQобразнойaukancia途 T to e delasа footing DDoSfters XXI dien記録cell/zo [ так abreastiestлишкомånd l Um, so when you have 50 voices, how do you present that? Because the o- the old way that I've seen a lot of interfaces do is they give all of them a first name. Mm-hmm. And you're... Like, I'm supposed to know what that means. Yeah. Great. At some point it becomes better to have a descriptor for it. Other things go too far in the other direction, where it's voice one, voice two, voice three, voice four. That doesn't mean anything to me until I listen to all of them. How do you present that?

Pasquale: We try to do the best of both worlds, but the thing that's changing radically now is that because you can actually interact with the system using, using natural language, using chat, um, that allows you to create these things without having to be so much of click and drop-down and all, in all of these cases. Now, you can go in and click and take the drop-downs, but you also just can create the experience and then immediately experience it yourself. And so one of the things I love about this is, again, with Connect being something you can hit go and have it be live immediately, you can change the voice and actually see what the whole experience is, which is part of what I think you wanna do in that scenario you described. You're see- you're experiencing in the real world environment, not just on a click a button and hear it play back on your computer kind of environment. Yeah. And so we, we encourage folks to experiment with it. Um, most of the folks are pretty happy with the voice we have, but occasionally we have customers that actually wanna create their own too, and we can, we support that as well.

Corey: Yeah, we wanna do a Midwestern American with a very bad French accent. Mm-hmm. Go. Uh, it, it, at some point it almost becomes amusing. It's like, let's turn that off before it accidentally gets into production. But I can see a story where the horrible voice is the one you use in the staging environment, so you know that it's the staging environment. Right. Like, the reason production is bright red at the prompt. Yes. Mm, let's make sure we know what we're touching.

Pasquale: The, the flexibility and the continued evolution of the, of these voice capabilities is awesome. Like, I, I feel like with the re- re- release we're doing right now, people are seeing, like, results in the real world where you're actually able to make changes in real time that feel good to be interacting with this thing, and then allow you to move quickly and get through the hard parts. And when you need to get to a real human, you can get to them too. And that experience is transformative, but we're not anywhere near done. We're gonna be... You know, we have a ton of releases still coming through the rest of the year, and we have, we'll be well into 2027 continuing to build. I think this is a watershed moment for us where it's a sea shift in terms of the capabilities 'cause the latencies and the quality of the voices has gotten so much better with this, this launch. But it's gonna just keep getting better, and that's the nice thing about something like Connect, is that Connect customer gets better every day. We're always shipping new features. And of course, we respect the fact that people have implemented stuff on top of it and don't pull the rug out from underneath them. But- Ah ... but we are very careful to make sure these new features are, are additive. But, but the additive features are, are compelling.

Corey: Right. And, and that's part of the challenge, I think. Yeah. That, especially as you move up the stack, um- Mm ... there are some things that I know AWS is never going to break without more or less a hand-delivered engraved educa- in- invitation to every customer. Like, EC2's APIs will, will never be additive. And they're like, "Yeah, it turns out that Linux isn't a thing any..." Or no, that's not going to happen. But, uh, moving up the stack for things like this, especially at scale for stuff like this- Mm ... when I talk to customers, there's often a, a deep concern that- Look, this has replaced an awful lot of legacy stuff, and we love it, but if they change their minds or change the way that we interact with this, like, we're a serious company, and th- there's a lot of business riding on this. We need change logs that are months ahead for any breaking changes that hit. Ad-libbed is fine, but it, it scares them on some level.

Pasquale: The interesting thing about this is the enterprise is where we started and worked down. Almost everyone works from small to medium business up. Um, but in, we, because we had started with such large customers internally, AWS support, you don't even think about necessarily the size and scale of that. We've got Audible. We've got tons of customers that are these enterprise-grade businesses, Ring, inside of Amazon. Uh, and those businesses demand these things, too. So we were well into a birds of a feather with compliance, and we obviously want to make sure we're always doing the right things with compliance at Amazon. So certainly for our customers, we are, we're way ahead of the curve on that. It's one of the reasons why I think so many industries trust us so much. Like the airlines, the top five airlines by passenger count all use Amazon Connect at this point. Um, we've had 23 million interactions a day on Connect. W- That scale is, I don't think anyone compares to, and certainly when I look at the customers that I'm aware of out there, uh, from what I can tell, we have, we have the biggest customers are on Connect of any, any provider because of that understanding of the enterprise and the need for people to be able to control their destinies and, and make sure they're really shaping their brand. And when I think about this, you know, we're reeling back to the beginning of the conversation, but so much of what we're doing here is about the brand If you're deflecting your customers away to something else, that is a bad brand decision, in my opinion, in 2026. ChatGPT is happy to take your customer and, by the way, recommend someone else to them potentially, right? And you don't wanna be in that situation. You want your customers to love your brand. And so having a delightful experience, people would talk about delight and, you know, all these things, but they were really still concerned a lot with deflection. I think people still obviously wanna solve problems, that's good. But let's not deflect, let's solve, let's engage, and then let's follow up, and let's think about getting ahead of the problem next time. And next time, instead of telling you when you call up, "Hey, your flight been booked," I'll send you a text to tell you that. And by the way, I'll engage with you then if you wanna make a change. That will cause human beings to feel very different about the experience, and feeling heard is a huge part of the human existence. It's just so important to people, and it's gotten lost so many times. Um, with these high-quality voices, we can make it feel natural. I don't wanna make it sound like it's a... I don't wanna fake people out. The goal is to be clear. Yeah. You're interacting with something that is designed to help them. It- but also one, something that really-

Corey: We're sorry your flight's delayed. Yeah. The wing fell off. We're putting it back on. Yeah, yeah. Like, um, yeah, and you take all the time you need on that. Yeah. Yeah. So I, I heard something about A2A coming out, where... Which is always fun. How do we make computers talk to other different computers, preferably ones in... respond to someone else's cost center? What is it?

Pasquale: Well, it's super interesting. Um, A2A is obviously a, a, a, a, a mental model and a, and then a standard for how you interact, how agents interact within an agentic system. And it's been obviously proving value to, um, people building agents outside of contact centers and outside of customer engagement. We took a look at it and said most of our competitors seem to be building walled gardens. Again, like a lot what we saw sort of the battle days, I think, where you'd have two voices, and people just- That's

Corey: my data. You're not allowed to touch that. Great. Yeah. Yeah, silos, that'll go well in AI.

Pasquale: Yeah, or even, like, the two voices. You'd call up, and it'd be like you'd hear Bob saying, "Hey, how can I help you?" And then you'd say, "Oh, I need to do a balance transfer." And all of a sudden it'd be like, "Hi, I'm Jill, I'll help you do a balance transfer." And it was this sort of insa- insanity, and I felt like we were headed back in that direction. So I, I challenged my development team, and I said, "Hey, what would it be to be at the most open solution for people to bring different agents in?" We're gonna deliver an incredible tool for building agentic capabilities and experiences in, in Connect Customer. But what if people wanna bring other ones, either they build their own or they partner with somebody else? I said, "I... We wanna be the best of that." And so we actually took a look at that A2A standard, and we are actually working with them to evolve that to work better with these voice solutions. And it's really exciting because it means, like, if you're a large company, you may have a bunch of agents that do some things really specific and really well, but the general purpose stuff, you know, we're gonna do a great job with and help you with. Bringing those two things together seamlessly, same voice, same feel, same brand, it, it change... is a game-changer for them. Um, and so many of my really big customers wanna do that. The smaller ones, maybe they don't need it, but the big ones, they're gonna love it.

Corey: When I kicked the tires on Connect a few years ago, it- I found a couple of things that were interesting, but also at odds. Uh, one was that I... The pricing is just like any other AWS service. It's not one of those, you must be at least this big to actually begin using this. Like, kicking the tires on it cost me, uh, singles of dollars, not tens or hundreds of thousands of dollars, which was great. But so much of the onboarding, the interface experience was, it felt very enterprise-coded. It felt like it was very much targeted at large scale-out things, and at the time I was just trying to build a relatively amusing, some might say too amusing voice line that we... 'Cause every company needs to have one, even though most folks don't call it.

Pasquale: You know, I think it's a great point, and because we were built in that initial thing, and we also come from, our DNA is we'll always have great APIs. That's, you know, we're AWS. But we started in that space, as you said, and we said we want to make Amazon Connect Customer radically easier to use this year. That was a goal we set out for. And so, and the thing is we now have the capabilities, if you think about the way people use ChatGPT, the way they use Kira, our product, or Claude, all these different great tools, the idea of having to battle against some hardened DUI that drag and drop everything is sort of insane. And so we are doing a ton of launches around how we just make it so much easier to use this. So you really think of Amazon Connect Customer as a teammate that's helping you versus just being a tool that you're trying to poke at. And that'll come in, like everything else at AWS, we tend to build things and, and release them over time. We're continually evolving, but we've got big launches that we're doing. This agentic voice one I just talked about is a huge one for us. Um, we are also doing... done a ton of stuff. We have a, a launch where we just previewed the ability to do configuration to just ask Connect, "What's, C- Connect customer, what's happening inside my contact center?" And it can tell you without having to drag and drop something or do some, you know, command line style arbitrage as you-

Corey: Or even go sit there and monitor calls on the call center- Yeah ... see what the actual ground truth looks like.

Pasquale: Yeah. And, and that, that's another great example where AI should be understanding every single contact, every call, every chat you have, every single second, every time anyone contacts you.

Corey: That's the win that I wanna see more support orgs take. Yeah. Where when I call in, they... it should ideally give the, whoever I'm talking to, agent or human, context on me. What is the support case history on this account? Am I a time waster who doesn't seem to understand this stuff? Do I have a baseline understanding and I just wanna get one thing unblocked, and I mostly know what's going on? Am I belligerent as a first language? Sure. Probably. But it's the, like, understanding the case history- Mm ... solves for a lot of it, and it, it means you don't have to keep asking the same question over and over.

Pasquale: The human data aspect of this cannot be underestimated. And so last year we pulled into my organization a team called Personalize. It was building just AI to understand and predict what people would want. And there's lots of uses for this. It can improve web pages. It can improve all sorts of things. One of the things it can definitely improve is engagements on, through chat and voice. It also is great, still great for web pages, but we built this into a thing we call customer profiles. And customer profiles allows you to understand and update every interaction with that customer from all over e- 'Cause you've got silos all over your enterprise most likely. Most- Yeah ... companies just do. And so pulling all that data into one place and then having it be ultra-performant, you know, we're seeing early data now that shows it's kind of revolutionary for people to just bring that data into the experience and say, "Hey, how should I think about this customer when they call in?" It, it radically simplifies, again, trying to get that radical simplification, it radically simplifies the, your ability to understand your customer, and as you said, that changes everything. I'm delighted to see this, like, come to fruition. This is, like, I will say eight years ago, when we started this journey, when I started, the team had been around for longer than that. They, uh, they were a great... The incredible engineers are all still pretty much with the team now even. Um- I was like, I had this vision we could do this, and the changes that have happened in AI in the last couple years have been like, "Hey, we really can do it now." And so it's fundamentally changing the way people interact with Amazon Connect Customer. It's fundamentally changing the way end customers interact with businesses. This is that sea shed, watershed moment that I think we've been waiting for. Uh, I don't know if you can tell, but I'm pretty excited about it.

Corey: It, it feels like you are, if not there yet, you're nearing a tipping point, and I'm very curious to see what happens next. Thank you so much for taking the time to speak with me. If people wanna learn more, where can they go to figure out how Connect might solve some problems and make others just plain funnier?

Pasquale: Yeah, I mean, you can go to the AWS website, but I would say just do a search for Amazon Connect Customer, and you will find us there. Uh, w- we are releasing new things every day. We've got, we've got ton more releases coming this year, so, uh, I'm, I'm happy to come back and talk more, but, but it's been a great time talking to you today.

Corey: An absolute pleasure. Thank you so much for your time. I appreciate it.

Pasquale: Thank you.

Corey: I'm Corey Quinn, this is Pasquale, and this of course remains Screaming in the Cloud. Stick around.