Transcript for "Get to know Recurly's agentic AI: Built for subscriptions": So I'll go ahead and start with the intros. So even though I'm the guy on the bottom of the screen, I'm gonna be your emcee for today. I am Patrick Rizzo, senior customer success manager over here at Recurly. And I have with me the one and only Niraj, Nick. I'm gonna go ahead and hand it over to him to introduce himself. We are. Thanks. both, by the way, out of our Greenfield, Colorado office, which is the best of the five, Recurly offices. Yeah. Thanks, Patrick. Hello, everyone. Niraj Nagg, VP of product. Been with Recurly for for a long time now, so I'm super excited to to be here and, walk you through some of the stuff that we're doing, around AI. Awesome. So a little bit about what we're gonna do here today. First off, this is part of our navigate program. I do wanna give just a brief introduction of what recurring navigate is. Navigate is our on demand customer success program here. So we do have on demand content, training content. You can self guide through learning paths, orchestrated by some of our wonderful colleagues here in the customer success department along with some of our excellent friends in product like Niraj, who help us create the content so that we can be the most useful for all of our merchants across all tiers from the smallest to small to the largest to large. We wanna make sure everyone gets the service that you want. So just a little bit of brief overview of what this is here. I do wanna say that we are live. So sometimes we might have a microphone problem or an Internet problem. We're human. I'm sorry, but at the same time, I'm glad I'm not AI. So we are live. You will get a recording at the end of this, and there will be q and a time. So please do post questions in the q and a. Up in your upper right, you've got a q and a there. Anything that's in the chat, we can always move over to q and a twos. Do feel free to engage. We love engagement from our merchants. This is how we grow. This is how we improve. This is how we are always getting better. So we will answer some of these questions live. Anything that doesn't get answered live, we will answer offline after. We will have your email address from your registration, and we will respond to everybody the best that we can. So without further ado, shall we begin to introduce our guests of honor, which are our compass agents, which is what we call our AI agents. Yeah. So. Thanks, we do. have So, so, you know, we firmly believe at Recurly and I you know, everyone's talking about AI. We've been there doing that now for the last couple of years now, and, and we firmly believe if you are gonna maximize the value out of Recurly, you know, understand how it's gonna help you drive growth, for your subscription business, you know, get recommendations and just use it as scale. We firmly believe from a product perspective, we have to provide that through, you know, AI assistance. And so we've embarked on that sort of like journey. We've launched, under our umbrella called Compass. Compass is basically essentially everything that we're doing that's AI related. We've launched a a few agents here to help you learn more about Recurly, do certain configurations, get the data that you need. We've, arranged those in a series of agents, and then those agents can execute different skills. We also have our Compass assistant. So if you're in the Recurly app and you see those little, you know, the the, the sparkly sort of icon on the bottom right of your screen, that's where our Compass assistant is. I encourage you guys to use that, ask questions, get answers. Again, this is all in the vein of helping you get to the things that you need in Recurly quicker and help you maximize the value of Recurly. So we wanna do some demos around some of the things that we've been doing. So the first demo that we wanna provide here is around our knowledge agent. So our knowledge agent is all about and we'll we'll talk about this in the demo too. Whoops. Oh, yeah. So we're the knowledge agent, is all about, like, understanding again, asking questions about, like, the the product itself. How do I do x? What are the best practices around y? Where do I go configure this? What do you mean by this? And this agent is is really strong in helping you, like, again, understand what Recurly is all about. So why don't we go ahead and queue up the demo? So if you come into Recurly, you can see we've got a compass area here, which is basically our umbrella sort of brand for all of our AI enabled features. Down here, you can see skills. You can see the different agents we have and the different skills they can execute. And then, of course, down here in our sparkly little icon here, if you click on the sparkles, we have our Compass assistant, which essentially is a chatbot that allows you to not only execute any of these agents and and skills, but also just ask ask ad hoc questions about best practices and and those sorts of things. So really excited to show you what we've been up to. I think what I'll start with is one of our agents. So let's say you're new to Recurly, maybe new to subscriptions, and you just wanna learn. Instead of going through and reading through all of our documentation or if you have a specific question about something, having to go, like, hunt and pack and find where that documentation exists. And then once you find the document page, having to go read the entire document page to get where you at, get what you need, we have a knowledge assistant here, and you can perform document searches here. So let's go ahead and try it. So you can kick off the skill and the agents just by clicking the, try it now or get started button. And, you know, I'm interested in understanding, you know, what is the difference between involuntary and voluntary churn. So I may be new to subscriptions. Just curious. It'll go through. It's gonna read all of our product documentation, our best practice documentation, our knowledge base, which is what our support team uses to answer questions that come in from our customers and pro prospects. So you can see here, it's done the research. It tells me what the difference between voluntary churn is and involuntary churn. Great. That's all great. But, like, how does how does Recurly specifically help me solve my involuntary churn issues. So now it has context of what you guys are talking about. It understands how Recurly use it what how Recurly works, what solutions it has to solve these problems. So it's gonna come back and tell you, hey. Here are all the different things that we can do to help you with, solving involuntary churn. So you can see, of course, there's account updater, intelligent retries. We have dunning management, you know, our our wallet feature, which is backup payment methods and so on and so forth. So now it's not only, like, just generically talking about involuntary and voluntary churn, but also it understands how recurly specifically can help solve those problems. So, again, encourage you to come in here, experiment with the knowledge agent, ask it questions, give us feedback. If it's not hitting the mark, we would love to hear from you. Thank you, Niraj. That was recorded, Niraj. We have live Niraj right here on the call too. Let's check-in on our q and a's. If anybody has any questions about the agents as you go, please do keep those in the q and a's, and we will answer a couple of questions in between each demo. Make sure that we're keeping everything relevant to everybody who is on the call here today. One thing I'll call out that I use the knowledge assistant for the knowledge agent for is oftentimes our internal Slack gets inundated with, hey. How do we define this or what's included in that? And I often go there for the documentation so that I'm not beating up our internal teams and fatiguing them with things that I can solve myself using this new agent. It leads me to the great places that I'm often able to share with a merchant and helps us solve your problems a lot quicker. So, Neeraj, it doesn't look like we have any questions here on the knowledge assistant, which tells me that you were so thorough in that video. Yeah. Let's learn about been another agent with column Roger Code. alright. The code agent. So here's another agent that we developed that, you know, if the knowledge agent, if you think about it, is more geared towards your admin sort of persona, the folks that are using the the admin interface on a on a daily basis. The code agent is really geared towards our developer persona, so folks that are integrating with Recurly, folks that need to understand, you know, what APIs to be hitting, how to, like, go do certain things. So, again, this code agent, it's pretty cool stuff. It generates sample codes and, you know, I'll talk a little bit less about it. Why don't we go ahead and and show the actual demo? Let's see. What other agents do we have here? We have the code agent. So just like let's say I'm I'm new to Recurly. I'm a, you know, someone who's responsible for growth. I get a lot of value out of the documentation search knowledge agent. But we have a lot of engineers and developers who use Recurly too who need help in understanding how to integrate, their, platforms with Recurly. So we have a code agent here which will help our developers and engineer customers, write code. So let's go ahead and try that now. And here, it's going to wanna get a little bit more contacts. You know? What kind of, coding language do I use? Am I Python? Am I Ruby? What am I specifically trying to accomplish? And, we can go in here. And I'm gonna go ahead and ask it to do something specific for me. So I'm gonna tell it, you know, I'm a a Python coder, and I wanna understand how to write a Python code snippet to go create a new subscription in Recurly using the Recurly API. So now it's not only looking at our product documentation, but now it's also looking at all of our API documentation, and looking at our developer kits and and all of those things that it would need access to, in order to provide a meaningful response, and that's exactly what it did here. So you can see here in Python, it's telling me how exactly I go and implement, a create subscription command through through the API. And you can ask it questions about anything that an engineer would wanna do. Maybe it wants to know about webhook event handling. So let's go ahead and ask it about that. And, again, it has context. It knows now that I'm a Python developer, so it'll show me how to set up webhook event handling for fill payments to the extent that it needs to be Python specific. It'll give me a response that's Python specific as you can see. Oops. It did exactly that. So here's all the the webhook event structures for failed payments and so on and so forth. So, again, this is a great sort of shortcut in understanding how our code works, how our APIs work. If you just need sample code snippets, you can quickly generate those through our assistant and through our agent. I love your examples because those are two of them that I use very frequently with my merchants. Another one that I often use is rolling out APMs. This is great for helping you write the code for that. I've had it. I'm gonna pick on digital wallets here. I have asked it to write me code in Python. I mean, you could pick your language, but I've said write me code in Python to offer Google Pay, and I use x y z gateway. And it's able to pull from our documentation and do that, making it much quicker for you, the merchant, to get new payment methods up and running without all the developer bottleneck that if you're anything like the rest of companies in the world, developer bottleneck is real. Niraj, we have a q and a here about our MCP, and I think this. is very relevant. Yeah. I'd love to to answer this one. Thanks, Felix, for posting that. So you guys can all see his question here. We we do have a a, plans to launch, our MCP publicly. More details on that in q four. So internally, we do use this MCP to generate a lot of the responses that you're seeing in the Compass Assistant. And so the vision there is we are investing more into making the conversations around, like, your data, your reporting a lot richer, as well to to really, like, you know, target that growth persona who has more advanced questions about, like, what's happening specifically with their, subscription program. So that is work that's underway right now. We're investing in there. And then we will also be looking at how we, can get better at answering questions around troubleshooting, you know, your integration. There's a bug you're seeing and so on and so forth. So, those sorts of things we're making investments into the MCP as well. And then once we got those investments in, we will launch the MCP, more formally for folks to, use in cloud, cursor, codex, whatever sort of, AI environment that you're using. So really excited about that one. And not to steal Patrick's thunder later, so there is a button up the top, sign me up for the q four road map. Yours truly plus my calling in product marketing. We will be going through the road map, where we'll also talk about some of the, additional investments that we're making in in MCP and and the AI stuff that, we're discussing today. So encourage you guys to attend that. Excellent. Thank you very much, Niraj. I'm going to, on the fly, name all of our agents' new names, and they're gonna get human names. So we do have another agent coming up here, Alfred k configuration. Great. Thanks. Patrick, I didn't know we were naming our agents, so this is fun for me to to hear these names too. So configuration or configuration agent, Alfred, I guess, is Patrick has affectionately named it, allows you to actually go execute configurations in your platform. So, again, when you're thinking about, I don't wanna have to go hunt and pack and figure out how to go create a plan. I can go ask this agent to go create plans and do configurations for you. And so why don't we see a a demo on how that works? And now let's say, you know, I wanna create a plan. I want some shortcuts around that. I don't wanna have to go, like, understand how to, like, configure a plan and go work through all of your documentation, go find the configuration page, click on a bunch of things. So I can go ahead and fire off the configuration agent, which is gonna help me create a plan. That's one of the skills it has. So I'm gonna tell it to do something, create a monthly plan for me. So go ahead and create a monthly plan, $99.09 a month with a fourteen day free trial. It's great. Now it's telling me, like, hey. You forgot some key information, so I'll go ahead and put that in there. And it's going to be, my, oops, my cool plan. That's gonna be the short code for it. So it understands that my cool plan is the name. A short code is my cool plan. It says it created it. Let's go see if it actually did. So if I come over here in my left hand nav, go into configurations, go into plans, and then I can hide this for a second and then go kick cool. You can see there's my cool plan, $9.09 in a month with a fourteen day free trial. And then I can open up the assistant again. I haven't lost context. So if there's modifications I wanted to make to that plan or create a new plan, I could easily do that without even leaving the context here. So, again, another way, AI assisted workflow to help you get to the actions and configurations that you need to quickly in Recurly. Excellent. Thank you very much. And you might notice that we are pulling from some of these skills in here. So we train the agents to have certain preconfigured skills as well that you'll find. So not everything has a skill, but many of our common uses have skills in there too. So any I wanna make sure we don't have any questions lingering in here. Jim, There is one gonna come oh, sorry. hey, Go. Patrick. Yeah. I did notice one question from Jim around, can code agent import a constant contact HTML email layout? There it is. Thank you for posting that. Sadly, there isn't a a way for the Code Agent to do that today, but this is why I love getting in front of customers as a product manager, hear feedback, hear the things that you know, the gaps that we need to, address. So this is great feedback that I can take back to the team. Again, we are investing heavily in this area, so, you know, we will continue to make enhancements, and I think this is a a great value add here. When you think about, the experiences we wanna enable in in Recurly, you know, the the AI, it shouldn't always have to be where you go to, like, Compass assistant. Right? So to Jim's point, we want the AI to be where you're actually doing your work. So imagine you are in the Dunning configuration window, and there's a button that says import your, you know, HTML here. And then we could kick off an agent that could go and pull that and then pull it in. Those are those types of experiences that we definitely want to to enable in the future. So so great feedback, Jim. Thanks for that. Awesome. I appreciate that. Do you have any questions on the configuration agent before we introduce our next agent? Alright. It's time for, I believe this is gonna be Elizabeth v data agents. Alright. Elizabeth, the data agent. So here is you know, we're just scratching the surface on this, but, a lot of our our users sometimes need access to data, quickly. And, you know, so here's an an agent and a skill that we developed to quickly run exports and get data out of the platform. So why don't we show a quick demo, and then maybe afterwards, I'll talk a little bit about where we're headed, next on on this particular Alright. Let's check out our, area. data agents. We have a data agent. One of the skills it can do is run a generate and export. So let's go ahead and try that now. And, really, the problem we're trying to solve here is, like, often you're asking, like, hey. I just want a dump of certain transactions or certain invoices. You know, I don't wanna bother someone. I don't know how to go get it out of Recurly. I don't know how to navigate the exports. You can see here there's multiple subscription exports, multiple adjustment exports. Really, I just wanna know, like, hey. You know, get me oops. Get me all of my past due invoices from the past thirty days. And our data agent's gonna run. It understands the schema of our exports, understands the context of my question. And it's gonna generate an accounts receivable report for me that has all of my past due invoices. So I'm gonna go ahead and say yes. I'm ready ready to regenerate. Go for it, please. And now it's going to run the export, and it should provide me a link that takes me to my export. So let's stand by. There you go. So there's the link. When I click on this, it takes me to our exports area of our application. You can see that the export is already run. It's ready to download. I haven't lost context of the conversation, and and now I have the access to it. So now this, again, is an easy way to go fire off and export without having to navigate, pick a bunch of different parameters on filters, and and things of that nature. So more assisted flows and workflows is is the game name of the game here. You know, I think as dads, we're not supposed to have favorite children, but the data agent is my favorite child that we have here on the agents. The data agent, I don't think a day goes by that I don't tap into it for some report for somebody. It's just data is useless if you can't use it. So why collect it if you can't put it in a digestible form? I love that you guys created this agent. I use it, like I said, daily. Niraj, I know you had some more you wanted to add about this as well. Yeah. And add wanted to days in there too if you have questions. yeah. Thanks. Yeah. So post your questions if you've got feedback. But, again, like, this is back to our earlier conversation around some of the investments we're making. So imagine a conversational interface where you can ask questions about your data specifically, how you're performing, get embedded sort of, like, charts and graphs, embedded, you know, tables that that show, what's happening with with your subscription business, being able to run custom sorts of reports and queries all through the Compass assistant. Those are the things that we're enabling through our MCP, which will then be available in the Compass assistant. For folks that have access to our, you know, advanced analytics, we also have conversational analytics available there too. So you can go and you there's a little, like, icon on the bottom right. You can click on that in, in any of our analytics views, and ask questions in a conversational way about any of your data, and we'll provide conversational responses. So so a lot of, like, great ways to, like, ask ad hoc questions and and get custom answers to your queries, that, you you know, saves you the time from having to, like, export data and hand roll your own sorts of reports. But but, again, join the the q four road map conversation that, that's coming, in October, and and we'll we'll explore that further. Perfect. We do have a question here from David as well. David is asking, do we have the I'll I'll I'll let you read that here. Yeah. Does the data agent include the capacity to schedule a report? An example would be the invoice detail. If I were able to run the invoice detail as of exactly midnight at the end of the month, then it could be used as an accurate month and AR balance that accountants and owners want instead of deriving it somehow. So, David, you can't do that today. But, like, the other feedback that we're getting, I will definitely this makes a lot of sense as far as, like, you need a certain cutoff on when your data is versus, like, exporting and then go create your own cutoff. So so that's something that, you know, I can share with the team and and see if we can, add some, enhancements around that. But that's great feedback and a and a great use case for for where we need to take this next. Awesome. I will, go ahead and plug that we do also have a revenue recognition module. So for those that need to use Recurly for additional features other than just subscription management, we do have revenue recognition as well. And our next agent, let's call this the double agent because this agent gets two videos. So without further ado, we have our retention agent. Yeah. So, you know, when we are out in the market talking to customers, talking to prospects. Everyone is talking about the challenge around retention. And, and so so we we hold that near and dear. And, like, as far as, like, our product strategy and our vision for the product to help you put subscription growth on autopilot, you know, again, through AI assisted tools. Retention is where we're really focused first on creating a lot of that automation. So, again, join the road map session on on, in October, and we'll talk a little bit more about that. But we've already made some inroads in in how we can help you solve some of your retention problems today. So without further ado, why don't we go and fire up the videos? Another one of our agents we have is our retention agent. That can execute a couple of different skills, one of which is helping you remediate involuntary term. And one of the best ways we found to, combat involuntary churn is by providing in app message to your customers or your subscribers that have a past due payment so they can go in and update their billing information, and then we can go ahead and retry that on a on a past due subscription. So here, the AI assistant is gonna walk me through setting up the prompt that I wanna show to my subscribers in my application. So your payment did not process as expected. The call to action is update now. I actually wanna change the background color, so I can go ahead and ask the agent to do that. And, of course, I picked the recurly yellow. I'm pretty happy with the way it looks, but I'm kinda curious to understand if I can do any AB testing. So I'm gonna go ahead and ask my assistant if, that's something that we can support. And it says that's a great idea. Of course, we do support that. You can adjust the title, the message, call to action, background color, any of those sorts of things. So it's good to know. But right now, I'm not gonna make any more changes. I'm not gonna set up an AB test. Now it's gonna take me to the next flow or the next part of the configuration here. So I need to go put, a link behind my CTA button so it takes subscribers to the place where they can go update their bubbling information. So I'll go ahead and then type that here. And now the AI agent is putting that link behind the button so that when there's a past due payment, subscriber will click update now. We get redirected to that link. Now you can see the agent's telling me here's exactly what you've set up. It's basically configured and ready to go. So I'm gonna go ahead and tell it to save my campaign. Campaign saved, and now it's walking me through. Okay. Now to actually make your campaign live, there's certain configurations you need to do in your app, which is basically put this HTML snippet, onto your app to trigger this. So this is a great way of of of one of our assistants walking us through some complex design, the complex configurations to get an involuntary turn prompt, launched and live on on your app. So this is a really exciting agent. Love your example. Love to see some of the great things that it does that I haven't even done with it. So this is why was doing the the demos. That and his voice is preferable to mine, so I apologize to everybody that's listening. But we do wanna make sure we get you the content, and I was readily available. So, I do wanna also call out while we're waiting for some q and a's that there will be a learning path available after this call, which does include videos, does include, really all the resources you'll need to be an expert, not quite near each level, but more Patrick level expert on each one of these agents and very well capable and equipped to use these yourself. We don't have any further questions yet in the q and a, but we do have a bonus video here on retention agent. Let's take a look at Recurly's retention agent. The retention agent is gonna be used for integrations to Recurly engage as well as optimizing our dunning settings. Let's get started by clicking get started under the optimized default dunning prompt. You'll see here it brings up our Compass assistant. Compass assistant will suggest a series of actions. We're going to be able to check our existing dunning length, review changes, and it can make the updates for us. Are you ready to get started? Oh, yes. We are. So we're gonna type yes, and now it's looking at our dunning length. It's telling us our default dunning length is ten days shorter than the recommended twenty seven days. Would we like to change it again? Yes. You'll notice here the agent is doing all the work. Shall we apply the change? I think you already know the theme at this point. Yes. So now we'll see it's checking off the tasks. It making all the changes to our done in campaign so that we can retain more of our revenue. I had to sneak in one of my videos in there too. We both came prepared, and, I had to sneak one of mine in. But literally five minutes before the call, before this webinar, I was actually doing exactly that skill with a merchant. So it is really fast. It's really easy to apply, and we wanna make sure that if there's something that has a best practice associated with it, we're able to make that easier on you and reduce some of that friction of implementing best practices. We do have one question here for Niraj. Is the retention agent available in Recurly Engage platform as well or just in subscription management? Great question, Katie. We are launching what we are calling, I think, the Design Studio AI agent, or assistant, I should say, in Engage, specifically. So if you are using our Engage product, and you go in and you are starting to design prompts, for your subscribers, be it a cancel save, involuntary churn, whatever prompt that you are creating, there will be an AI agent that assists you in designing that prompt. The other thing that we've done to help with prompt creation is we do have a a sync, with Figma. So that's also an AI based sort of, like, sync. So if you are doing your designs for your prompts in Figma, we can use that, that feature to automatically sync in the designs from from your prompts. So, so lots of cool things that we're doing in Engage and is is, and, again, with our with our focus on, like, helping you drive subscription growth, helping you drive retention, you know, the the more we can do in our engaged product to help you guys engage with your subscribers the right way to drive that growth and retention, the better. So so we're excited about the advancements and enhancements we're making there. I don't wanna give too much of a tease, but I did get to play with the engage agent a bit. And it it makes even me look very CSS skilled. And CSS is not my strong point. We'll just say that. So we have one more demo available here. One more agent, which let's call him, overarching captain compass. Yeah. And so, you know, all the demo videos have shown, like, you go to an agent, you execute a skill. All of that is just available in our Compass assistance. So you can think about those those agents, those skills as, like, shortcuts to do specific things. But, really, you can execute a lot of different actions and a lot of different sorts of, conversations within the Compass assistance itself. So why don't we go ahead and and see a a demo of that? Alright. So that's a a look at our agents. And like I said before, any of these agents, any of the skills that they can run are all available through our Compass assistant. And you can ask Compass assistant anything. Like, one of the things that I like to do is ask about best practices. So I want to launch a new Dunning campaign. What are some best practices I should follow? This is a question that we commonly get from our customers. It's, we even run workshops with our customers on how to, like, create effective dunning campaigns. But now we have, you know, some assistance here from our our, AI agents. And you can see it's given a pretty thorough response around oops. Still responding here. But, around, you know, how to optimize your dining window length, twenty seven days if it's a monthly plan, what you should do with your email communications, the cadences, the personalization, the tone of them, you know, highly recommending segmenting your Dunning campaigns and so on and so forth. So this assistant is great in not only just executing actions and giving you access to data, but also when you wanna understand what some best practices are on the things that you're doing. Try to run those through our Compass assistant, instead of having to, you know, log a support ticket or pick up the phone and call your customer success rep. Instead, you can, you know, get some insights and glean some from in some information from our our Compass assistant here. I feel like we couldn't have queued up our our missed question any better and, Shannon, I apologize for that. Shannon, you basically you have earned your way into my heart right away here because while we're still on the call, you are already going into action. So a plus work. I'm giving Shannon an a plus, but we have a question here. I just tried to optimize Dunning and it suggested a length of 27, but that's really for monthly subscriptions, which, Niraj did just mention in the video and not for annual subscriptions. One of the issues with the documentation we've found is recurly documentation is mostly focused on monthly and not annual subs. So this is some feedback for you, Niraj, and anything you would like to add. Yeah. This is great feedback, Shannon. And, so in q four, we will be launching in beta mode a new sort of recommendations engine. And what's different about that compared to what we do today is that our recommendations will actually look at your configurations, look at your plans, and then provide tailor recommendations based on your configuration. So for instance, if, we recognize that you have an annual plan, the best practice that would be, recommended would be a sixty day dining window, not a twenty seven day window. So so this is great feedback, very aligned to to where we're headed, and it makes a lot of sense. You need recommendations that are tailored to your business, tailored to your configurations, and that that's exactly, the path that we're headed on. Thank you very much on that. Do we have any other questions about any of the agents as we start to wrap things up here? We do have some more time available for q and a, and I wanna make sure as we have that time, we make ourselves available to answer that live should we still be here. So, couple call outs while you guys ask those last couple of questions is, again, I wanna call attention to the fact that there is a learning path available after the call. I do recommend go through it. It's really educational. We worked hard on making it. We have worked hard hard on making those things useful, valuable to you. Also, if you have questions that come up after the call, we do have some office hours available where several of us CSMs make ourselves available, and we are available to answer those questions. We have another webinar coming up here, road map and vision. So it give you an opportunity to sign up for those as well. And I wanna call out if there are no other questions, wanna call out we always appreciate a review. So if you are finding it valuable enough that you joined us today, first off, thank you. And secondly, please consider leaving us a review on g two. We do use that to reach new customers. We're always looking at, you know, what's next. That's that's been our motto this year, get ahead of what's next. So we wanna make sure we're hearing your feedback, and we are listening to you. Hence, Neeraj and I are here right now. Yeah. Please also, as Jennifer had mentioned here in the chat, we do want feedback on these learning paths too. We wanna make sure that they are the most useful thing that we can do for you. So we please please communicate with us. We're here. We listen. Thank you very much for your time today. We're gonna give a couple more minutes in case any questions do trickle in. But I can't say enough, thank you guys for joining us. I know you had a lot of things you could possibly be doing in this hour of your day, and to join us really is humbling that you would choose to spend your time this way. Alright. I'm gonna take that as we have no questions. But if anybody has any better names for the agents, I don't think I did the best job. I just did a an on the fly naming job. Hope you had fun. Hope you enjoyed joining us. Thank you again. Thanks, everyone, and thanks for the feedback. It's super helpful. Oh, we we do have one more question. Can the agent create a plan or 20 currencies, Niraj? Can they create a plan for 20 currencies? I don't know off the top. Again, so this is you know, it's a nondeterministic way. You know, we're working with the LLMs here. So, Gilbert, what I would say is give it a give it a try. You know, start maybe simple, like, hey. Create a plan with, you know, in 20 US dollars. And then as a follow-up, can you add, you know, a $10, euros, or whatever it is that you want and see and see how it does. Let's let's challenge challenge the agent. And, and if it's not doing what you expect or not what you want, you know, we are logging, all of the queries that folks are asking and where they're running into to problems, which we use as input to inform the next round of enhancements that we make to the back end MCP. So, Gilbert, I I'd say give it a whirl. Give it a try. See what happens. And with that, I believe that's a wrap. Thank you, everybody.