Tom Morgan: 0:34

it's time for another ACP and it's a special one this time because this is reinvent week. And so we have a, a whole deluge of things to talk about. There's lots and lots to get through on this episode. Luckily, I've got Alex with me to help. Hello, Alex.

Alex Baker: 0:50

Hi Tom. How's it going?

Tom Morgan: 0:52

Yep. Very good. Thank you. Just kind of trying to pick our way through all the new stuff. So let's set the scene. Reinvent was two days ago as we record this And a whole bunch of announcements, obviously across all of AWS it's, it's a w S's big event for announcing new things. I think it's fair to say Amazon Connect got a fairly big chunk this time round of, of new stuff and new announcements, right?

Alex Baker: 1:17

Yeah, definitely. Yeah. Like you say, quite the, quite the deluge of, of new stuff, wasn't it?

Tom Morgan: 1:22

Hmm. It really was it's, it's an interesting dynamic like sometimes you get, busy years and quiet years and I think this definitely felt like a, a busy year and it, I think it sort of makes sense a bit where if you look at what. Some of the announcements are, and we're gonna dig into them. There are definitely some themes and some key areas. We've tried to watch through as much as possible and we may miss some stuff. But I think we've picked up what we think are kind of the, the important things we. Pasqua de Mayo is the vice president of Amazon Connect, and in his session he actually said there were 29 new features for Amazon Connect. Which is kind of a lot and but he split'em across four key areas. Action with ai, elevating your workforce, transforming data into relationships and accelerating outcomes. And so those are kind of like the, the broad things. And he described this set of releases and, you know, this set of features as Amazon Connect, kind of bravely moving into the agent world. He said, actually what was interesting, he said you could now, if you wanted to, you could run Amazon Connect with no human agents completely autonomously. He said, you, you won't wanna do that. That's not a good idea, but you can do it. In the same way that you can still do it a hundred percent. You know, completely human. No agentic at all. I think the right spot is somewhere in between those two things. But there's a whole bunch of features now to kind of help you do that. And so we, we'll go through them. Where should we start? Alex? Is there a sensible kind of starting place? I've got a couple of thoughts around like the age stuff we could, we could sort of start and go through,

Alex Baker: 2:53

Yeah, I, I think, yeah, it's probably just gonna be sort of throwing a, a dart and seeing where we, we hit, isn't it, it's interesting though, Pasquale's comment there about being able to run, connect completely agentless, isn't it? It's sort of bringing, bringing a bit more substance to the thing that lots of people are saying about how. Contact centers and customer experience in general are quite, ripe for for yeah, removing live agents.

Tom Morgan: 3:17

So more agent stuff in Connect. So there's been a, a bunch of things, features and functionality to kind of help with some of that. Should we just go down the list and, and call them out, Alex, like, is there an order

Alex Baker: 3:28

for, are you looking at a list and I'm also looking at a list. Have you

Tom Morgan: 3:32

We've, I think we might, let's have different lists. We've got slightly different lists, but let's compare and contrast, like do you wanna do a top three from your list? And then I'll see if I've got anything to add.

Alex Baker: 3:41

Well let, let's do top three sort of in order of, of reading anyway. So, and I, I'll call out. The, the list I'm looking at is from an an tonk from from AWS who has a, a good LinkedIn, Amazon Connect, what's new page? And he's just put in a single post what looks like all of the, the launches from. Reinvent and then links to the documentation. So that's quite a good place to have it centralized. That was one thing we were saying earlier, is that there's just, there's so much information, so many posts from different sources, different people, different sites. It's quite difficult to, to keep a track. So this was, this was quite a good centralized version. The first one on that list is Amazon Connect now provides native testing and simulation capabilities which is quite a cool concept. I, I recall sort of back in, back in my days of working on Cisco, Cisco had quite a nice call simulator where you, you could, you could just press a button to place a virtual call and you'd then see it sort of run through all of your, your contact flow, your your call flows and, and use it for troubleshooting. See what, what route it takes when you, you enter certain values in IVRs, that kind of thing. So I'm, I'm guessing this is, this is probably gonna be a nicely up to date version of, of something like that, but lemme see if I can find some more,

Tom Morgan: 4:58

Yeah, that makes sense And, and super useful nowadays as well with, you know, as we're gonna talk about some of the agent stuff where it's not just like the IDRs and the call flows, you're testing, you are potentially testing an entire call flow, like an entire end-to-end call process of, you know, invoking the right agent, getting the right answer, the wrap up, you know, everything that happens after it. So yeah, super useful, very

Alex Baker: 5:21

Yeah, de definitely. Here we go. So I found some notes that this, apparently, this is in preview. So one thing I will say, I have had a look through the, the console. For some of these new re releases, and quite a lot of them I'm not seeing yet. And I, I guess that's probably a mixture of, some of them are in, in preview, like this one seems to be, and then some of them are, are probably in the process of being rolled out as we speak.'cause you know that they are, I acknowledge that they're quite, they're quite fresh announcements. But yes. So as a result, haven't sort of had a. Play with any of these. So the, the notes we have on it, Amazon Connect now allows you to test and simulate contact center experiences in just a few clicks, making it easy to validate workflows, self-service, voice interactions on your point there and their outcomes. So for each test, you can configure test parameters, including a caller's phone number or customer profile, the reason for the call, the expected responses, and business conditions like any after hours scenarios or full call queues. After executing the tests, the results show success or failure based on your defined criteria, along with the path taken by the simulated interaction. Plus you get detailed logs to, to diagnose any issues. So from that, it definitely sounds pretty comprehensive, doesn't it? It's, it probably a bit more than what I, I mentioned at that sort of the sort of call tester. It sounds like you can do the, the self-service interactions as well, which you'd expect, you know, going down the, the route of. More ai powered interactions. But yeah, sounds pretty cool.

Tom Morgan: 6:54

Yeah, it really does. It's very good. It comes at a good time. I was, I was kind of wondering whether we should build something like this ourselves as like a, you know, either just for the company or as like a community thing. But I'm, so I'm kind of glad that that AWS had done it for us'cause it's, it was getting to the point where it's like, we really need something like this to, to kind of test the end to end. So, yeah. Brilliant.

Alex Baker: 7:14

Okay. What have we got up next? So we've got Amazon Connect launches, automated email responses using conditional keywords and phrases. Let's see if we can find some more on that one. This one, this one doesn't specify preview at the start, so I dunno if that means that it will just kind of drop in and become available. The notes on this, say Amazon Connect now allows you to automate email responses and agent routing logic using keyword and phrase conditions, helping organizations increase self-service, reduce manual handling time, and improve routing accuracy. For example, if a customer sends an email asking if a certain product is in stock or is checking on their shipments status, an automated response can be without involving agent.

Tom Morgan: 8:00

Okay. That's pretty cool. That, that sort of rounds out, interestingly, rounds out the email part of Amazon Connect and actually extends it out from being. More traditional what you think might think of as a contact center and to make much more omnichannel building some of those capabilities. It's, it's an interesting move, a, a expansion move, I suppose. Adding those, adding those features.

Alex Baker: 8:22

Yeah, it'd be interesting to dig into that.'cause I, I know already you, you had the ability to, to send sort of templated emails as an auto response. So I, I guess this is perhaps going more into that reading the body of the incoming email and categorizing it and responding it based on the, the body, which Yeah, we, we've, we've built some stuff like that. Built it from scratch, custom development. But it's, it's nice to see that it's it's coming into the product natively Now,

Tom Morgan: 8:48

Yeah, definitely

Alex Baker: 8:49

Amazon Connect Chat now supports agent initiated workflows is the next one.

Tom Morgan: 8:55

Okay, so that's interesting. So like a proactive outbound or proactive kind of chat.

Alex Baker: 8:59

Yeah, so this apparently enables agents to send interactive forms to collect sensitive data or share policies and disclosures within customer chat conversations. So, for example, the customer needs to update their address. Agents can now send a form that customers complete without leaving the chat interface.

Tom Morgan: 9:17

Oh, nice, nice. Yeah, some of the stuff that was being talked about in some sessions is, is about trying to reduce the amount of context switching that agents have to do on calls. So anything that could stop'em, having to go into another system, come back again and try and keep them in that kind of single pane of glass is, is good and, and better for the agent experience. So that's good.

Alex Baker: 9:38

Yeah, definitely. It sounds like a nice add on to the, the step by step guides type functionality where you're sort of guiding agents through a particular process flow. But yeah, more involving the, the customer, obviously.

Tom Morgan: 9:51

Yep. Yep. Okay. I'm gonna do a couple on my own, give you a break. And so mine are focused a little bit more around, I suppose developer things, but that's natural I guess. But connect launches a model context protocol support, MTP support, which is kind of really interesting for agents. And we talk about some of that agent stuff. So outta the box, there's gonna be tools for Common tasks. So things like updating contact attributes and retrieving case information. So they're gonna sort of provide their own MTP stuff as well. And then you'll also be able to provide your own MTP servers. And so Amazon Connect will then be able to see them and that will enable their AI agents to do that kind of customer self-service. But also the employee assistance side of things as well to help. Provide more information and help complete actions. So this is kind of really moving from just providing information to actually doing things as well. And I think that's, that's really, really interesting.

Alex Baker: 10:48

Yeah, that's really exciting, isn't it? For I guess both sort of agent. Agent backed and also self-service may be where you, you have more complex workflows that, like you say, you can, rather than just providing a bit of information, you can have it go off and perform different actions. An example on the notes here, an AI agent can automatically look up, order status, process refunds, and update customer records during self-service interaction without requiring human intervention. That really kind of takes self-service to that, that next level, doesn't it?

Tom Morgan: 11:20

Hmm. Yeah, it does. Yeah. No, it's, it's good. Yeah, so that's gonna be out the box for doing some common things within Connect, but also you'll be able to provide your own tooling as well. Which you'll do kind of through either flow modules you write there, or you can also integrate it with agent Core, so Amazon Bedrock, agent Core. Don't think we're gonna have time to cover'em here. There's a bunch of new announcements for Agent Core as well, so if that sounds interesting. This is probably not the podcast to cover them, but go check them out because there's a whole bunch of AI related announcements. And it. In these announcements and in some of the stuff that's coming out about doing this, this starts to make sense now in where stuff should happen. So we know, agent Core is AWS's method or their way to, to make agents and, that's a. Nice sort of developer experience for not just writing and making the agents, but testing them, evaluating them, and doing all those sorts of things. And so it's good to see that there's a hook now interconnect. So these two tools can talk to each other. So you haven't got to try and sort of choose between two or, have a lesser experience for your agent building and, and build less agents because you can't use agent Core. The two are kind of coming together, which is really good.

Alex Baker: 12:27

Yeah, agreed. Yeah. The other, other thing just to pick out of the notes on that one I did see, so this includes some updates to how flow modules are used and yeah, you, you, you touched on it, but using flow modules as MCP tools, so kind of bringing a bit more power to your contact flow modules by the sound of it.

Tom Morgan: 12:48

Okay, cool.

Alex Baker: 12:49

to see how that materializes.

Tom Morgan: 12:51

Yeah, and sort of related, but kind of related realtime chat message streaming which I think is gonna be a big one for making the conversations that people have with agents just feel a little bit more natural. And so, this is, realtime streaming for AI powered chat interactions essentially. That's good. Like it makes that sort of more natural natural sort of thing and, and gives some confidence I think that somebody's actually doing something. So it's a bit like you see the sort of typing thing you know, when you are in your messaging platform. And then you can sort of see the responses appear in the notes they say they, they're gonna appear progressively. So we'll see what that looks like. I dunno if you've seen this, I'm not quite sure how this is gonna show up, whether it's something that is something the developers need to lean into and enable or work differently in some way or whether it's just gonna happen. I.

Alex Baker: 13:41

Yeah. So you, you thinking, this is, I guess when you interact with something like a chat GPT you're kind of, you, you have that wait while it, while it thinks about it and processes its response, whereas this might, it, it's sort of streaming the response. So you're not, you're not sat there waiting for, for as long. Is that what you, you

Tom Morgan: 14:00

I, I, yeah, exactly that. I think and what I'm not sure about is whether developers will be able to do the same thing. So for instance, whether, I dunno, you'll be able to. As, as you receive data from some stream, whether you'll be able to kind of pipe it essentially in real time. And then it will kind of appear and essentially replace. And then it is a, I see it similar slightly to if you've ever done something to get the, like the transcription. Out of event bridge, something like that, you get partial sentences until the whole sentence is there. I, I see it a little bit like that, where you can sort of send and then replace and replace, you know, as, as things improve or as you get more information. And so you are, you're kind of throwing the, the data to the customer as quickly as you get it, basically. So yeah, I'm, I'm hoping that's how, that's how that works.

Alex Baker: 14:43

Yeah, there, there's quite a lot of these, as we read through them, are really sort of gearing up towards that, that next generation of, of self service, aren't they?

Tom Morgan: 14:52

Yeah, definitely. And actually we should also talk about the big noisy el elephant in the room which is Nova. So, so that's, that's in here as an announcement, but it's kind of one that I know lots of people have been talking about and are interested in. So like top level, Nova, so is, it's an LLM but it's, it's for voice essentially, and it, it provides that nice just like we were talking about actually with chat message streaming Nova Voice does voice message streaming so it feels much more natural. And so. That is now coming to connect. Right. Do we know much of the details about that yet? Or is it still kind of in preview or do we know it It is not got preview tag on it, I don't think.

Alex Baker: 15:32

it's not. It's not, and I saw it I think it was mentioned that it was available in the two, main US regions east and west. But no, nothing about it being in preview as far as I've seen. Again, I haven't seen it. Anything, sort of, anything that says no for suddenly pop into the, the console. So maybe it's, it's coming in the next week or so. A couple of things in the notes for it. Amazon Connect. Introducing self-service capabilities that enable AI agents to understand reason and take action across voice and messaging channels to automate routine and complex customer service tasks. Skipping down a bit with integration from advanced speech models from Amazon, Novas, voice self-service experiences now deliver more natural and adaptive interactions. And a couple of, a couple of things on that'cause. We at Cloud Interac have been sort of playing with, with Nova, so a bit, one of our colleagues, Bentley has, has done a couple of really good demos and showing some of the, the key benefits to, to Novas. So, so how, because you, you mentioned it's, it's sort of voice to voice it doesn't have to transcribe somebody's voice. Process it as text and then come back and regenerate the voice. So you, in theory, you're kind of cutting out a bit of latency there, hopefully. But also in, Bentley's demo, which is well worth a look. I think it's on our YouTube channel. We'll put a link in in the, comments. But it's, it's really good at sort of matching, the customer's pace and, tone more than something like Lex ever was. And it's really good as, as well at, at elegantly handling interruptions and barges from the customer.

Tom Morgan: 17:12

I think we're gonna try and insert a bit of Bentley's clip now and then we'll hopefully if it all works, that's what you should hear about now.

Nova Sonic: 17:21

Hi there. Welcome to Cloud Interact Insurance. How can I assist you today? Whether you need help with your policy, have questions about coverage or need to file a claim, I'm here to help.

Bentley Paton: 17:31

Uh, yeah, could you just get my policy details for me please?

Nova Sonic: 17:34

Of course. To get your policy details, I'll need your full name and your policy number. Can you please provide those for me?

Bentley Paton: 17:41

Yep. Uh, my policy number is POL 6 5, 4, 3, 2, 1, and my name is John Smith.

Nova Sonic: 17:48

Thank you, John. Here are the details for your policy number POL 6 5 4 3 2 1. Your policy type is homeowners and it started on November 15th, 2021 and will end on November 15th, 2024. If you need any specific information or have further questions, feel free to ask.

Bentley Paton: 18:08

Uh, could you gimme some information about the type of cover that I've got, please?

Nova Sonic: 18:13

Sure thing, John. Your policy includes the following base coverage, dwelling coverage, up to the value of your property, personal belongings up to a hundred thousand dollars, liability coverage up to$300,000, and loss of use benefits. Additionally, your policy has two enhancements, storm cover, which adds coverage for storm related events. And temporary relocation cover, which extends standard temporary housing. Housing by covering additional expenses for hotel stays or short-term rentals, and provides a daily living allowance during longer displacement. If you have any questions about these details or need further assistance, feel free to ask.

Bentley Paton: 18:50

Uh, could you tell me what email you've got on record for me, please?

Nova Sonic: 18:54

Of course, John. The email address we have on record for you is john.Smith@example.com. If you need any updates or have further questions, feel free to let me know.

Bentley Paton: 19:03

And just one more thing, uh, could you tell me when my last claim was?

Nova Sonic: 19:07

Certainly, John, your last claim was for an amount of$12,000 and it was filed on April 25th, 2025. The status of this claim is currently open. If you need any updates or have further questions about your claim, feel free to ask.

Bentley Paton: 19:21

That's great. Thank you.

Nova Sonic: 19:23

You're welcome, John. If you ever need assistance again, don't hesitate to reach out. Have a great day.

Tom Morgan: 19:29

So as you can hear that's gonna make a, a big difference. I think that's gonna be game changing for some of this other stuff that's coming in as well. And it's all coming in at the same time, which is why it feels like so many announcements. What is so many announcements, but also why it feels so exciting, I think because you're having this new way of. Customers being able to interact with the agents, but you are also getting a whole bunch more capabilities about what those agents can do when it comes to the MCP server lookup. And then,, increase, integration into connect So the two things together I think are gonna make a a real big difference.

Alex Baker: 20:02

Yeah, definitely. Yeah, super excited about that. And people have been, people have been talking about Nova, so, and the integration with Connect for quite a while now. So yeah, really excited to, to see how it's set up in the, the console and everything.

Tom Morgan: 20:15

Can't wait to can't wait to get started with that. There's also some more AI stuff when it comes to agent summarization as well. I think,

Alex Baker: 20:22

I certainly got one around AI powered case summaries. Which was, which was new. I, I think that was one that was sort of, it was, it popped out, maybe the, night before reinvent.'cause I think there were a bunch that came out. Then.

Tom Morgan: 20:34

The one I was thinking of is actually better integration with agent force. Agent force service from Salesforce. So that's, that's like a real time agent assistance and, and contact summarization. So if you use Salesforce, go check that out because that sounds like it's gonna really help. Get recommendations. Yeah, enhance monitoring for supervisors and then automatic post contact summarization to help agents update Salesforce cases. Yeah, that sounds really good.

Alex Baker: 21:04

Yeah, that does sound good. We've definitely got a couple of customers that, that use Salesforce that will be pleased about that. The one that I was talking about, AI powered summaries for cases. So where for a while Now obviously we've had the, the gen AI summaries for, voice calls and for chats. We now have it for cases as well. So quick scan over the, the notes. Reduce manual wrap up work, provide complete context into, into customer issues, help resolve cases faster so agents can generate a concise case summary, even when the case spans multiple interactions, which is quite nice, where you've got that on ongoing case where somebody's called in, emailed, maybe come in via a chat as well. Being able to summarize the whole set of interactions is really powerful.

Tom Morgan: 21:51

That's cool. Yeah. Yeah. There are. A few kind of similar or related announcements just around analytics and reporting I think there's some stuff on self-serve conversational analytics I saw as well. But also more for automated performance evaluations, which I know we've spoken about before.

Alex Baker: 22:09

Yeah, I, I guess both of those things, it's sort of back to that, we were talking about if you are, if you're bringing in more and more automation in, in place of live agents, obviously you want the same level of visibility into what those AI agents are doing. As you have on your, your live agents. So yeah, it makes total sense that you'd want to run the, the performance evaluations on your AI agents and be able to, to report in depth on, on what's going on in those interactions.

Tom Morgan: 22:36

Yeah, that's really, that's, that's, I think that helps with, to your point as well, it helps with the trust and helps people adopt these tools if they can have confidence that they're working well. One of the other things that is in preview is ai, iPad predictive insights, which I think is interesting. So this is built on top of profiles, which I know you just talked about a little bit, I think. But I think is subtly different, possibly. So you, yes. I think, did you mention case summaries?

Alex Baker: 23:02

Yeah. And you're right, it's, it is, it is a, its own, its own separate point and announcement. So yeah, definitely worth touching on this one.

Tom Morgan: 23:09

Yeah, so this is quite clever. So it builds on profiles and comes with five recommendation algorithms to analyze customer behavior and interaction history and therefore provide information to agents or potentially, I suppose something that you could maybe use in a more automated way. So the five recommendation algorithms are recommended for you. So obviously based on their history, based on who they are similar items frequently paired items popular items. And trending. So that's really cool, I think that's really interesting to see how that works and, and see some of the information, some of the data and that comes out of that stuff like that's really interesting.

Alex Baker: 23:48

Yeah, it sounds like those two kind of work quite well in. Conjunction maybe. So yeah, the, the sort of looking back over a whole case and all of its interactions and then providing predictive insights based on all that and, and maybe a load of other data about what might be a, a good thing to, to suggest next upsell opportunities, maybe, that kind of thing.

Tom Morgan: 24:08

And, and actually it reminded me of something that Pasquale said in his session as well, just'cause they, they got into talking about, big places that use Amazon Connect and they've got lots of big partners, obviously but they also use it internally. So Amazon customer service using a hundred thousand customer service agents. But for their audible, contact center and, and, and that area, they, they match up agents they route to agents based on the shared reading history of the users. So you're more likely to be go to an agent that reads similar books as you, which I thought was really cool. Like apparently it just, it may means that the interactions between agent and customer is just better, like the which kind of makes sense when you say it, but it's really interesting to see them do that use data in that way.

Alex Baker: 24:49

Definitely. Yeah, that does sound really cool, doesn't it?

Tom Morgan: 24:52

Yes. Right. What have we missed out? What's important that we didn't cover?

Alex Baker: 24:57

I was gonna mention and I suppose maybe'cause this was a, a slight bugbear of mine working with, with knowledge bases in, in Amazon Qing Connect, but there is one there's another point around Amazon Connect now supporting multiple knowledge bases and integrating with your Amazon Bedrock knowledge basis. It seemed when we, we've been working with it in the past that you could kind of set up a single knowledge base, which Queing Connect and some of you, maybe you're queing connect in, in Lex intense could all, all integrate with, but it didn't sort of have a degree of separation very easily. So when we've, when we've been doing customer demos, for example if we were doing a demo to a public sector organization. The data was all sort of in the same knowledge base as then if we did a demo for a healthcare organization, it's all sort of in there and, it didn't seem like it was that easily segmented. So I'm hoping this is going to make that process a bit easier where you can have one of your lines of business having its own specific knowledge base and it's a bit more kind of kept aside from, your other lines of business or, or parts of the organization.

Tom Morgan: 26:06

Mm. That's cool. That's cool. I'm just looking through the notes as well. It's like that, that's definitely a, a, a useful use case. I think the, the other one as well they mentioned is being able to search in parallel. You, if you do have data in a few different places doing that search in parallel to, to get back information from across different places and aggregate it as well. So that's, that's interesting.

Alex Baker: 26:26

Yeah, agreed. And it might, it might help as well thinking about sort of guardrails. So, if you have a particular bot, being able to make sure it sources all of its information from a particular knowledge base or area of a knowledge base so that you're not getting hallucination or, or it, it's not providing information that it shouldn't be based on the context of what it's doing.

Tom Morgan: 26:51

We are gonna try and keep roughly to time, so, let's see if there's anything else that we should have talked about. Any big things. We definitely haven't covered all of them. Like I say, 27 new features. And like we've definitely spent more than a minute talking about each one, so we haven't covered them all. I'm just looking down to see if there's any really big ones. I've just seen one which is support for third party speech to text and sex speech AI model as

Alex Baker: 27:13

Yeah. Good one. Yeah.

Tom Morgan: 27:14

There's some really good models out there and not, you know, the built-in ones are good and getting better as well. And it'll be really interesting to see how how they work with Nova so as well and how that works. But yes, you'll now be able to integrate deep gram for speech to text and 11 labs for text to speech. It's a bit unclear from reading the the, the fairly short announcement blog post about it, whether it's any third party speech providers or whether it's just Deep Graham and 11 Labs. But yes, that it's, it's good anyway to see that being opened up, if you like, as a, as an alternative for people.

Alex Baker: 27:49

It makes me wonder if there are more, more to come. And, and I guess for, for those people that might have invested already in, in other sort of AI speech providers. Outside of Connect and who are maybe moving to Connect, that might be a sweetener in that they, they then, they don't have to kind of totally tear out what they already did on, on Deep Gram or 11 Labs. They can now move to connect and, and continue to use what they've what they've invested in there.

Tom Morgan: 28:18

Yeah, absolutely.

Alex Baker: 28:20

The only the other one I was going to, to touch on is Amazon Connecting Hunt is agent assistance capabilities. So it would be, it would be nice to, to see what. What that looks like. Provides customer service representatives with new AI agents that guide them through customer interactions by recommending actions, retrieving information, executing tasks on their behalf. So, yeah, kind of touched on it perhaps around the, the, the conversation we had on, on the MCP servers. But it, yeah, it sounds like it's sort of supercharging, the queing connect where you had the, it would in real time, it would surface knowledge base articles. It sounds like this takes it a bit to the next level and you're doing stuff like that, that sort of full process automation or, or assistance.

Tom Morgan: 29:05

Got it. Yeah. Which is an important part of the story actually. When we say agent, we tend to think about self-service. But actually an equally important part of that is assisting human agents. Very, important part of the process and as if you can increase productivity there, but also just make it a better experience for agents as well. It's, it's better for everyone. So,

Alex Baker: 29:24

Yeah, definitely. I can't remember where I saw it, but somebody was talking about aiming to sort of make your existing agents superhuman by, by providing them with, you know, so much of this, of this you know, really useful AI assistance in the background. I.

Tom Morgan: 29:37

Mm. Yeah, definitely. It really feels thinking about all of these features that are coming and or have arrived and are, you know, rolling out now. It really feels like quite a big step for Connect right now. I think it, it was an important, I think an important step as well. You know, I think they needed to, like the, the world is talking about age agentic behavior, and particularly around the self-service side of things. You're sort of starting to see solutions that, would solve some of that problem. And then only transfer to. To connect if you needed to talk to a person. And so I think it was an important move for them, I think, to get in front of that and, and provide these tools. But equally, it, it means that they can offer that capability for agents as well, which is, which is really strong, I think. And yeah, makes it a, a really compelling sort of platform change, I think. And yeah, it feels like a, feels like a big, big enhancement or round.

Alex Baker: 30:26

Absolutely, and it makes that, you know, the announcement about unlimited AI pricing, which we covered ages back in,, it makes it sound like there's a lot of sense in it, right? So there's so much more of of the AI tooling coming into it. You can see why that was, why that was rolled out the way it was.

Tom Morgan: 30:45

Yeah, absolutely. It's gonna, yes, it, it, it's making more and more sense and it just doesn't, it feels like something that is so deeply ingrained that it will, yeah. Having something like that's unlimited could really make sense and yeah, you have to wonder if that was the long game when it, when it first came out.

Alex Baker: 30:58

Yeah, agreed.

Tom Morgan: 30:59

Cool. What a, what a what a foray, what a romp. Through the the feature list that was, thank you ever so much for helping me understand it. There's loads and loads of other things we didn't talk about. There's a whole bunch of. Of AI stuff. So if that's interesting to you, go check out the keynote because there's lots and lots in that. There's AI coming to other places as well, not just Amazon Connect. Some really interesting stuff around Quick Suite as well, which I've been digging into. But for now I think it's time to bring this episode to an end. Thank you very much, Alex, and thank you for listening. Be sure to subscribe in your favorite podcast player. That way you won't miss the next episode. Whilst you're there, we would love it if you would rate and review us. If you have colleagues you think would benefit from this content, then please let them know. To find out more about how Cloud interact can help you on your contact center journey, visit cloud interact.io. We are wrapping this call up now and we'll connect with you next time.