Name: Get BFCM-Ready with AI Agents
Uploaded: 2026-09-22T01:01:09.517Z
Duration: 41 min 48 s
Description: Get BFCM-Ready with AI Agents

Transcript for "Get BFCM-Ready with AI Agents": Good morning, everyone, and, thanks for joining the, webinar series that we continue in in APAC. If you've joined previously, you'll know that we've been doing this webinar series focused on your AI voice agent for the last couple of months. We've had one with our Salesforce integration. We've had one with your HubSpot integration. And this time, we're gonna focus on the ecommerce industry, and we're gonna have Quinn speak to that. But I'm Frank Eagleton. I'm the managing director for Asia Pacific, in for ECCO. I'm based in our Sydney office. If you haven't been to our Sydney office, we're in Blythe Street. We've got about 40 people in the office. We'd love to meet you if you're ever in town, so please do take the opportunity if you have it. Like I mentioned, today, we're gonna focus we kind of thought with where we are in the season, where we are on the year with Black Friday and Cyber Monday, it's, it's probably a good time to to talk about how we add value in the ecommerce industry. And I know that, Trent has had a lot of fun with building a couple of agents that he's gonna show you guys in the next couple of minutes as well. So, I think we're gonna get a lot out of this webinar today. Before I hand over to Trento, a couple of things just to call out as well. So there will be a recording that will be shared with everybody after this meeting. So, you don't have to worry about taking notes or anything like that. Focus on what Quinn's saying. It's gonna get a lot of value from that. And then if you do wanna book a meeting, have a chat with one of your account managers, there is a book a meeting on the top of the screen. So you just book that, and we'll collect the details and have a meeting with you. And as per normal, there will normally be some questions. Feel free to enter the questions on the side of the chat. We have people that will answer that as well. But at the end of the day, we will also have some time left for you to have live questions with, Will, who's our head of, customer success based in Sydney as well. So he'll be monitoring the chat and facilitating the live chat at the end. With that, I'm gonna hand over to Trent who's our forward deploy engineer, and, he'll take over from me. Enjoy. Okay. Very nice to meet everybody on the call. Just make sure a thumbs up. People can see me or hear me. Yep. K. Good. I'm seeing a nod from a couple of people on the call. Hopefully. Good. Alright. Awesome. Well, very nice to meet you all. My name is Trent Hall. I'm the forward deployed engineer here in our Sydney office for Aircall. My job is to make sure that, basically, when it comes to preparing our customers for whatever, time of year it is, I guess, for supporting their customers through the use of the AI messaging and VoiceSuite on the Aircall platform, that's my day to day job. So, yeah, work with lots of customers sending these things up, going live. And I've also, like Frank mentioned, has done this webinar a few times before. I've gotten a bit of feedback, from previous webinars that people wanna see stuff happening and some demos and stuff like that. So today is a bit more interactive. I will be trying to make a few calls to our AI voice agents, and I also have a messaging agent available to go through today. So we just, so I pivoted on the last one, to give you an idea of what both voice and messaging AI agents look like on the Aircall platform. So one of the things that, is coming up for the retail industry, obviously, is the, big Black Friday Cyber Monday series. But, ultimately, what that Black Friday and Cyber Monday essentially do is they very much, open up already existing kind of pain points in retail customers. So even though that it is a vastly larger time of year for volume when it comes to people asking where their orders are, especially towards Christmas time, I guess submitting things that are wrong, etcetera. Ultimately, for most of the time throughout the year, most retail companies I speak to already have a scenario where they're missing calls or potentially, trying to cater for volumes, and Black Friday and Cyber Monday just really exposes that for the worst sometimes. So when it comes to how to scale, when it comes to catering towards, I guess, people that would be looking to extend their teams without having to hire tons of people. Hiring in itself for most retail organizations is a tough one, so we can look at how we can use, I guess, some automation through AI voice and messaging agents to address that. So, not gonna really go into too many more slides. I do wanna dive actually straight into the platform to give you an eye and a look into what that is going on. Now, obviously, on today's call, we might have a mix of people who are currently using Aircall or who are not, but that's okay. So just for the people that are not currently using Aircall extensively in their businesses, this is what the dashboard in Aircall looks like. Basically, we have all the bits and bobs on the left hand side that go through everything from integrating with your systems like Shopify or Gorgias or Zendesk or HubSpot or whatever you're using alongside of your primary, order management system. We also have the ability to do things like manage users in here and numbers for distributing air call around your organization. We'll look at what a call flow looks like and how easy they are to set up. And lastly, we do have an, couple of really cool features in here as well. For those currently on the air call platform, we've just added Ask Aircall. We're coming out with innovations all the time. This is like a, like, an agentic layer to the Aircall platform where you can ask it questions and things like that. And it actually goes a long way in terms of setting up AI agents, which we will talk to in a sec. Other than that, live monitoring, analytics, and all those other things I won't be covering today, but we'll be diving headfirst into our AI voice agents. And to navigate to that, it's as simple as clicking the button on the left hand side, and you do have the AI voice agents and the AI messaging agents tab. Now just a heads up, if you have never set up an AI voice agent before, you do have fifty three minutes for every, AirCall customer to play around with each month. That goes a fair way when it comes to testing and setting up. And then the other part is if you haven't ever set up an AI voice agent before, you do get a handful. I think it's about hundred extra minutes at the beginning of the month for you to do your initial setup. So that is available for every AirCall customer. Some of the, some of you on the today's call have already purchased bundles of minutes. So any questions about the commercials of how to bundle up the costs and things like that, again, deflect to your account managers, and we can go from there. But not to delay in today's webinar, I do wanna run through what an AI voice agent setup looks like, just at a high level before we dive into a few test calls. So we do have a couple of agents called Sol today. It's It's pretty gender neutral name. It could be Sully Sullivan or it could be, Solana or something like that. So, bear with me. But as we can see, when you dive into the agent builder itself on the left hand side, we've got configure, deploy, and monitor. And then from there, we are immediately, as we jump into the platform on the part where we give it a name, a description, and its voice. So feel free to if you do give this a try after today to give it a name. I think a lot of organizations do. Gives it a bit of an identity and make sure you pop a description in there. But when it comes to the voice, we have quite expanded quite a number of different voices that we have available on the platform itself. You'll be hearing, obviously, Ella today and Cooper, I believe, that is the other one. But there are a few others out there, and we we do support custom voices as well. Couple of extra things, filler phrases, background noises. You know, this is, all these extra configuration parts that we'll we won't dive into every one of them today. Because if you aren't aware, this time slot, most Tuesdays, I hold an agent building webinar where every week, I will, go through the beginning to end of how to build an agent. But as for today, we're gonna look at essentially some of the superpowers that agent has that go a bit further than those average bills to assist you during these times. Because let's face it, when it comes to the building of agents, they can do a lot by having a good knowledge base, and they can have a lot of call scripting. But it goes a a lot, lot further when we look at use cases where it interacts with your system. So the first thing that I'll mention is what you can see here on the left hand side, which is called an AI action. Now we have tried the best we can to make it as easy as possible for you to connect to your ecommerce platforms and your support platforms across this time because most of the time when it comes to retrieving information, it's a pretty stock standard process. One One of the most popular ones we see in the ecommerce realm is the get order details from the Shopify connector. So as you can see, all it takes for me is to add a few clicks in here. And once that I click through, I can add that action directly to my agent, and it now gives that agent the ability to go and look up orders inside of Aircore oh, sorry, inside of Shopify. So this is the first step that we're seeing today in being able to give your agents what they need to go that one step further to hopefully resolve calls by themselves when it comes to, when it comes to actually servicing customers. So getting order details is obviously a neat thing that it can go and do from Shopify. But when it comes to doing a lot more, which in terms of might be, in terms of might be touching other parts of the Shopify platform, we can build lots and lots of actions to do that. So here's a few custom ones that I'll show you for today. So for example, once it retrieve those orders, it will actually go ahead and offer to send a tracking link to the customer when they ask it to. So this just uses the, internal API inside of, AirCall to be able to initiate a text message to the customer. So when it comes to, building the agents themselves, before we jump into what this one list sounds like, one of the biggest things is, you know, how do we get it to do exactly what we need it to do. So there are two things to keep in mind about our agents is that they're very adherent to what you can see on the screen here, which is called its call script. Now its call script is designed for you to be able to basically give it simple instructions as to how that it should carry the call, what that it does over the course of the call, and in this case, what it will do not only with the with conversing with the customer on the call, but also with touching other systems that it's connected to. So you can see here search customer by phone, when the caller connects, and then getting the order details comes in there. So, basically, over the course of the call script, you can just simply declare when it needs to use its different tools throughout the conversation to add value to, I guess, and bring data in so that the calls become relevant and personalized to the person that is calling in. So there's no magic trick here. Basically, it's just simply writing in paragraphs when it comes to actually running through each one of these steps. If you do if the agent, I guess, goes through a certain cadence that you don't particularly like, all it takes is for you to come in here and write an extra statement. So we'll see today how that it collects emails and, I guess, how to edit that. But, ultimately, 99% of the way that agents configured falls on this conversational guidelines. We actually have an amazing amount of articles and use cases online right now that you can go out and have a look at different examples of how to write these conversational guidelines with copyable scripts. But if you do have any questions around how that I built these today, feel free to reach out to your account managers, and they can grab the examples from today's webinar to pass over to you for your first builds of maybe something that connects here. So like I said before, this one is just using Shopify. It is going to reach out to Shopify to grab order details, and it is going to be able to give you some information about that order back. In my case here, what that we have is my record inside of Shopify. It looks like this. I have a pending payment order in 01/21/1982, and I have a couple of other orders that are basically, I think that they've been fulfilled. So what we can do is we can call this agent up and see if he can give us some information about, I guess, essentially, my track record as a customer with this Shopify store. Alright. So let's let's just go back here, and, hopefully, this is nice and clear for everybody. Thanks for calling Solstice Supply. I can help with your order status, tracking, or stock availability. What do you need? I was just looking to follow-up on my order, see where it's at. One moment, please. Sure. I can help with that. If you have the order number from your confirmation, go ahead and give me that, or you can give me the email on the account. Yeah. My order number is 12182. Just a moment. Thanks. I've got order 12182. It's still pending, and it has not shipped yet. Is there anything else I can help with? No. I was just checking on the status of that one. Thanks for that, though. Thanks for calling Solstice Supply. Take care. Okay. Cool. So that is, you know, I guess, a good demonstration. Did everybody hear that? I can see some people leaning in. Was that okay? Yeah. Good. Alright. Awesome. So when it comes to, gathering information around that, the orders supplying that information online should be pretty simple. And hopefully, you're hopefully, it's a nice, voice as well. I know that our voices are sounding quite a bit more natural recently, so that can be quite good. Now the other thing that we do with, I guess, a little bit of an extra step here as well is have the availability of not just searching by phone, but giving the opportunity to actually run these at Call Connect. So the useful thing and the one bit of information that every caller has when they they call in is the ability to actually use the phone number that they're calling in on to actually go ahead and look up that order quickly ahead of time so that we can actually start to gather the information about what that is going on on that call, or what is going on with those orders before the person gives us the actual order number itself. So there are a couple of ways to see it in these actions. You can see here that we can use search number by order, but then straightaway go and get their recent orders. And, if they have a question about their recent orders, also send a tracking SMS. So we can have another go at this in a second and a call in from a number that's on the actual, Shopify order itself, and we should be able to grab, I guess, the most recent orders from the, from Shopify. So in that regard, Shopify has an amazing suite of different APIs that the agent can interact with. A few of the other use cases that I've seen from customers so far, customers have asked about things like, for example, can the can the agent look up stock volumes? Can the agency, fulfillment statuses? Can it see things like whether or not that that person has claimed a discount coupon? The answer is yes. So basically when it as you might have noticed so far, when it comes down to the agent being able to retrieve information or actually look up information from across the the core life cycle, it completely depends on what's available via API. And for most of the ecommerce providers that we've integrated with, we do have an open API interface to allow basically grabbing that information by using things like the phone number that they've called in on, using things like the email that they've supplied potentially, and that goes a long way in being able to grab the information that we need to be able to talk through the phone call itself. So let's instead activate that and hear what it's like when we can collect the information, immediately on CallConnect, and then we can at least ask what our current orders are without necessarily having to confirm both our emails or order numbers, stuff like that. Press 1 for orders or 2 for support. Thanks for calling Solstice Supply. I can help with your order status, tracking, or stock availability. What do you need? Yeah. I just wanted to understand what my, what my recent orders were. One moment, please. Sure. I can help with that. Before I share any order details, can can you give me the order number from your confirmation or the email on the account? Yeah. My email is Trent, trent,.Hall,hall,@Aircall,aircall,.io. Thanks. I heard trent.hall@aircool.io. Is that correct? No. That's wrong. Let me read it back to you. It's, Trent, trent,.hall@aircall.io. So that's a I r c a l l dot I o. Got it. Just to confirm, trent.hall@aircall.io. Is that right? Yeah. That's correct. Your recent orders are number 12182, status pending, then number 12172, status paid, and number 12162, status paid. Can you give me a information about the tracking on that second order? I can help with that, but I need the order number you want me to check. Is it number 12172? Yeah. That's the one. One moment, please. For order 12172, it paid and fulfilled, but I don't have a tracking link for it. Is there anything else I can help with? No. That's okay. As long as it's being fulfilled, that's cool. Thank you. Okay. Cool. So interesting one there. And I deliberately left it in in that way for today's demonstration because a lot of people are like, hey. You know? Can we get it to read back the email faster? And that brings me back to something earlier that I said there. So 90 like I said, 99% of the configuration of these calls and how that you want them to play out is very literally in here as well. So when it comes to recognising the caller or asking for an email, you can see here, that I've gone ahead and asked the email itself. And all that it takes for me to do in here is say, please, when, repeating the email, read back everything everything before the at symbol letter by letter, and then naturally read back everything after the at symbol. Right. So that's all that it would take in that case for it to instead go, is it t r e n t dot h a l l, and then it would say @aircall.io rather than spelling that out. So hopefully that gives you a look and feel of a couple of the different types of calls, how that it can interact with grabbing different details from Shopify. And in this case, we can make a small tweak to speed up the way that it collects information or even add statements in there, like, read it back the first time. If you get it wrong, then spell it out incrementally slower, all of those kinds of things. So hopefully, that's a good understanding of how that we can hook up to a data source and instead grab details and supply that to customers that might wanna get updates or are calling in or something like that. Now, obviously, when it comes to customers calling in and the next thing that you're probably likely to say is, great. They're probably calling up because their fulfillment status hasn't progressed or they're looking to understand where it is and then they're gonna wanna complain. So the next most popular thing that we often see are support use cases. So when it comes to the support use cases, most of the customers that are supporting on, that have Shopify with us, have Gorgias. So the answer is yes. We do support with a Gorgias system. In that regard, we also can interact with these APIs to do some really cool stuff there as well. Now if this is a demonstration of Gorgias, but please keep in mind that we have really strong integrations with most ticketing platforms like Zendesk and HubSpot as well. So lots of different configurations of ecommerce that are paired obviously with some sort of support tool, have the ability of integrating in. So this one here is a little bit different. So this case, we're going to extend on our ability to grab the order details from, from Shopify itself. But in that case, we're gonna use the, use the ability to grab the details from Shopify to then search Gorgias to actually be able to see where that a ticket is at. So just a reflection inside of Gorgias, this is what my customer, record looks like in there. You can see that we have a, I have a ticket inside of wrong sized order inside of Gorgias, which gives me the ability to, to call in and potentially update this case with a call. So this is me. I've lodged a issue with a wrong sizing. So let's have a here of what that it might look like instead of a just a simple follow-up with an order status, but what that it might actually be like when you're calling in for a support use case. Press 1 for all of the support. Thanks for calling Solstice Supply. How can I help today? Yeah. No. I just had an issue that I was trying to get in contact with somebody about. One moment, please. Hi. This is Sol at Solstice Supply. I can see you have an open case about getting the wrong size on order one two one six two. Is that what you're calling about today? No. Actually, it was, I actually got a different order that was just delivered, but the package was ripped open, and the, the product was damaged inside. I'm sorry about that. What happened with the damage in a few words? Yeah. It looks like that it might have happened during transit, and it was basically torn open, and the product was wet and damaged. I'm sorry. That's really frustrating. Do you have the order number for that one? I don't. That's okay. What's the best email for the team to follow-up with you? Cool. So you can see there that in that case, what we've done is, you know, use, I guess, some of the data that exists inside of, inside of Gorgias to potentially just give the customer the option to say, yep. This is what I was calling about or I wanted to get an update. And in that case, we can go ahead and use that system and look what's currently going on. This gives us a really good chance of allowing customers to reach out and update what's going on, if they've posted the package for a return order, if they needed to follow-up with something as well, like they wanted to cancel the return or they wanna keep the product or things like that. In that case, one of the biggest things that we see across some of our e customer base ecommerce customer base is that in the in this in this scenario where customers are reaching out and wanting to leave information, what they can as well do is leave comments on the ticket and the agent can give them updates backwards from what was left on the ticket, giving you more of a conduit where customers are leaving and updating a system of record, giving you an auditable trail in terms of whenever a customer might call back. They can call back as often as they want to receive those updates via tickets, or they might be able to submit additional tickets in, above and beyond what they already have. One of the things that we've seen often is that when ecommerce companies give the opportunity for a customer to say to a customer, oh, is this about a previous ticket? It does reduce that repeat or duplicate ticket volumes, especially if they're done online maybe, and then they call up, and then they get a confirmation that the ticket they submitted from a different channel is there and it's in the queue. So it definitely does have, the capabilities there of supporting not only from a, from a order management side, but also a support side so that you're, I guess, internally in your organizations over these really busy periods, it might become more of a scenario where that, you might put your agents front and center of the support line, take those tickets, and follow-up as they go along and solve them incrementally online rather than consistently having to pick up the phone and reprioritizing everything that those people are doing as they respond to more and more phone calls. So one of the things that, I do often want to show is what it looks like when deploying these. So today, I was just using a standard IVR where I could press one to demonstrate the order and two for the agent itself. But one of the, one of the, I guess, the good things about orientating your agents into a call flow is that you do have the power of the you do have the ability to have lots of different, options in terms of creating your call flows inside of Aircall. If you've never seen these before, this is basically at the top. This is when a call comes in. So you heard that today when I was calling this number, and as it progresses through, that's when the call ends. So in this case, the standard IVR is what we normally see from, in this case, lots of businesses are putting their press 1 for this or 2 for that. But often as organizations evolve, what they can what they may wanna do is actually extend the capability of that routing to be done by an AI voice agent instead. So in this case, we can go ahead and have an agent actually listen to what the customer is wanting to ask about. It can try and solve questions before that it progresses on and then can, escalate the call down to a further couple of agents if it needs to. So in that case, a lot of organizations put a receptionist agent in there with all their return policies, with their promotional information, and things like that. So when customers call up, they can ask the questions, and the agent can maybe deflect a few questions before they actually progress through to individuals that are having to pick up that phone line. So we roughly say on average about 20 to 30% of calls being able to be deflected at that initial stage to slow down the overall volume that might be further down when hitting a person. Now other things as well when it comes to how the organizations feel comfortable in deploying these is that your support, agent itself might actually sit behind human beings. So in the case where you're obviously striving to provide exceptional customer service, you can, of course, take as many phone calls as you can, but you're naturally going to miss a few occasionally. So agents can sit behind teams or individuals to be able to take those overflows when the call volumes get exceptionally high. And then the last most common one that we see as well is the ability to actually add a time rule in here so that you can cater towards one certain flow happening in hours and then maybe a specific agent being dropped in here after hours that can, in this case, update tickets, give us information, maybe look up order statuses, and those kinds of things, which goes a lot long a lot lot longer way than just a voice mail that might just be collecting, you know, voice mails over the period of offline, which which you're gonna have to come back to the next day and listen to lots of recordings or look through lots of transcripts. So hopefully, that makes sense, and it hopefully, you can see that it can be quite easy to navigate through a very visual interface in terms of deploying whatever it is between your teams, between your agents, and between any option that you find useful. And, of course, any of it can be done and published and saved, once that you're I guess, in this case, when an agent or when things are configured in the way that you need to. So that's all ready to go now. As soon as you click publish, that phone line will be taking in during business hours this way and not during the other way. And, of course, you can start to play around and use quite a number of different tools you have available here across your experience to create everything from audio messages through to ring via API, which will give the opportunity for the agent to route the phone call maybe to the case owner or to the lead owner if you're in sales motion there as well. Now lucky last for the last five minutes before we jump into some questions. I do wanna mention one last thing, which is, of course, our messaging agent. So the messaging agent itself is a relatively simple process of very similar steps in terms of getting started, giving it a name, allowing it to speak in certain languages, and then from there, setting up your scripting to allow the agent to work in a particular way. So when it comes to the so when it comes to the actual, agent itself, it's very similar. It's actually a little maybe a bit more simple than the, than the voice agent in terms of how much information you need to give it, but it does have an exceptionally large knowledge base. So for this one here, what, what that I've hopefully touched on is that in the case of some of maybe your loyalty style, programs, you might have something like this in your marketing platform. So this is a workflow inside of HubSpot where once the customers during the Black Friday or Cyber Monday, if they're in a loyalty tier that's important to your business, we have native tools inside of your workflows and automation systems to be able to send those SMSes, to customers that might prompt them with a unique code and things like that. So in that case, it's a bit of a surprise and delight, but you may want to craft a message that might elicit a response. Right? In that case, you might say something, reply here anytime if you like a hand in finding something, or it might come in terms of, delivering or how to actually execute on the loyalty offer. So one of the really cool examples that I've seen recently is a AI messaging agent that helps, customers kind of find useful product recommendations alongside what they've already purchased. So in that case, it actually goes, let me know if you need a hand with finding something. And in that case, what that the what the inbound messaging people will start to converse with the agent about is what, what how the the promotion works or what products it might be associated to. So in this case, all it would take is for you to actually deploy this onto a phone line, and it's sitting there. It does have the ability to look back in previous messaging history, so that people can converse with it and just ask, I'm, wanting to know how that I can use this promo. So in that case, the they might have received a message triggered out from your system. But once that they receive that message, it might come in reply something like this on where that it would, whether it would actually reply and start to go back and forth to coordinate with the in this case, in the customer, where that it could potentially use those discount codes or when that they might be available. So all that, all that it does is pivot that kinda AI generated or AI agentic experience from not just voice, but also into messaging so that you might not necessarily want your customers talking to an agent, but rather messaging it back and forth. And the good thing is that when this escalates, I guess, people can jump in over the top and coordinate and go back and forth with customers as well in the air core workspace. So the messaging agent itself just basically gives you opportunity to deploy, whatever that you'd like to, from a messaging lens, which also goes a long way in additional use cases for support as well. One of the more common examples that we see on the messaging agent side is actually the ability for the messaging agent to coordinate a return. So in that case, the messaging agent can see pictures, which is kinda cool that we've released recently. It can coordinate the process of somebody and walking somebody through the process of a return. So going to the right place, submitting the right information, filling out the right forms, or processing the order through potentially a certain link that they need to with your ecommerce platform. And then that gives a guided ability for the messaging agent to coordinate that over a longer period of time. So, you know, for often for things like warranty claims or support requests, there might be multiple steps needed, like go and look up the serial number. Here's where the serial number is for that particular product. And in that case, then include that in this field, in this form, and then we will get that off to to submit that under warranty for you. So, yeah, lots of opportunities to walk customers more slowly through the process because you're not having to kinda deliver that in real time, giving customers a slower mechanism to follow-up on things that might take a little bit more time depending upon your products and your customers. Cool. So that all in all gives you at least a bit of an intro to some of the use cases, some of the deployments, and some of the, some of the ways that ecommerce companies are scaling and covering, covering, I guess, the gaps in their organization that they can't fill with raw people power, but and hopefully gives you some inspiration as to what that might look like today, in deploying an AI voice agent across your ecommerce organization. So I will hold up there. Hopefully, that's been insightful. Hopefully, that's given you some examples as verbatim and things like that. We're about ten minutes left. So I do wanna hand back over to my colleague, Will, who will be coordinating, I guess, any questions that might be coming through. And if you do have any, just keep in mind on the right hand side, you have the q and a part where you can type in and ask questions. But other than that, hopefully, that's been a good example of a few use cases. Thanks, Trent. We actually do have a few questions here. I'll just go through the list. How does it handle an angry customer, and does that re does it recognize kind of a sentiment trigger? Yeah. It's a really good question. So maybe on that support agent, you actually heard the agent be a little bit sympathetic. Now that was kind of some of the ways that I added into the conversational guidelines for it to work. So there are ways that I've seen customers add in things like, you know, be considerate during a process where it's a broken order or something like that. Maybe change the sentiment if it's, you know, if it's a customer asking about different things, but those can be included in the conversational guidelines. The last option that you have as well is that agents don't only just have to look after things a 100% by themselves. We can turn on what's called, the ability for the agent to escalate. So in the scenario where somebody doesn't wanna speak to an AI agent, whether they want, support from a person, whether or not there's a particular maybe there's a loyalty tier that you don't want the agent handling. You wanna just put it through to an individual team to support them. Those escalation branches are there for, for the agent itself to escalate calls to the human beings. Thanks. And then kind of in the same conversation, there's a few questions about how does it handle certain languages. Can you speak to kind of the process behind that and how we're updating those languages specifically for Australia and New Zealand customers? Yeah. So when it comes to language support, I guess, organizations that are servicing customers internationally from Australia and New Zealand might need the support of an agent that is in a different language. We do support over 20 different languages now. So when it comes to deploying those agents, it is one agent per, I guess, per language. So what it would take is for you to duplicate the agent itself and switch over the language and then make it available as is. A lot of organizations choose to have a voice agent, to have an IVR that would say, would you like support in different languages or assigning a regionally specific number. So for example, if you sell into, let's just say, Southeast Asia and you wanted to have a specific number for that region, we can help you with deploying that inside of AirCall to attach your, regionally language, language aligned agent onto that specific number. So it's a more intrinsic method of contacting your company. Great. There's a question about a real a realistic go live time considering that, these peak seasons are coming up and coming up quickly. So in your opinion, the customers you've worked with, what have you seen the go live timeline look like, from a discovery call to launching? Yeah. So some of the more common examples of agents that are deployed for, like, getting order details or things like that, because that's essentially there's not much variance between ecommerce platforms and the ability to do that, the deployment schedules are actually very, very quick. So having an agent looking up details, supplying it back to customers. Because that we've done it so many times before, it makes it really easy for us to deploy those agents very quickly within a two to three week period. For more of a tailored, use case where it might coordinate with a certain system or have multiple calls, so for example, look up on Shopify, interact with Zendesk or or Gorgias for support and things like that. Often, what that we would do is sit down and have a workshop, in the first week, do a build and a, I guess, approved concept with you where that you would then have the chance to call it, iterate with us, and have it to a point where you might distribute amongst your organization for further testing. Once you're happy for that process, once the agent is kinda refined at the two to three week mark, we then maybe roll out for a soft launch with your customers where that you can distribute maybe an hour of traffic a day or a couple of hours or, out of hours for your agent to take calls, giving us the ability for a softer distribution to your customers and see how that they react, see how they respond, and see how the success of the agent would go. That case, there's normally an iteration round there as well, which represents about the three to four week mark. And then at that point onwards, we kind of let it loose in the wild, with our organizations of which we'll start taking a significant volume of their calls. And in that case, we'll do another further check-in just to make sure it's doing everything that it's supposed to. But at that point, it's generally kind of a set and forget, especially considering a lot of the agents in the ecommerce brands have a pretty narrow scope of responsibility. So finding and supplying orders, updating and locating tickets, supplying, you know, I guess, logistical information, stuff like that. So in that case, normally, they're pretty once that they've set up and the organization is happy with it at about the four to five week mark, then it's pretty much, you know, like I said, set and forget. So anywhere from a three to six week process, majority of the time, I would say, is that if you don't have things like credential access to setting up and connecting the agent to things like Shopify or Gorgias, make sure you coordinate with your IT teams ahead of time, and that normally isn't a hurdle, but it can be held up if, as you're stepping into this that you don't align amongst the team in terms of getting what you need. Most of the time, that needs list, we can forward over to you ahead of time. We do have a one pager that can help you get prepared for deploying an agent with all of what I just said in a nice handy PDF. Yeah. And I think it's also good to know because I know we have a few customers, on the webinar that have existing, agents that are live, that aren't potentially working with you, Trent. So in that case, that really you've you've done the brunt work. So we work with our bundle customers not only on the creation and rollout of these agents, but also the optimization of them. So similar to what Trent was saying, if you have an existing agent, but you wanna tweak it and you really just wanna optimize it to make it the best version of itself, working with Trent to be able to do that through multiple iterations, and then also be able to expand the amount of agents that you actually have on the platform that are interacting with your customers. Yeah. For sure. Look. There's lots of opportunities, when it comes to deploying agents to take take the load over this busy time. But, ultimately, one of the most common ones that we at least see, if anybody is, you know, a little nervous about dipping their toe into these kind of things, you know, naturally, if you do have an out of hours time slot where you don't provide support or you don't provide, I guess, an active ability to answer customers, it's a really great place to start. Because at the end of the day, if customers are just hitting a voice mail, I'm sure you can probably do a little bit more than that or at least allow the agent to give you an actionable output in a structured format the next day rather than plowing through voice mails, especially as we get to Christmas. Great. Gilmore, I saw your question about the account management. I'll reach out to you after this. And that's all the questions that we got, Trent. So thank you everyone that who attended. As a reminder, please, if you wanna meet with your account manager to chat through next steps, click the book a meeting button. You'll have the recording sent to you after that to dive into it in more detail, and always feel free to reach out to myself, if you need some direction on where to go next. But thank you everyone for joining, and have a great day.