Name: 3 Strategies to Elevate Your Travel and Expense Management with AI
Uploaded: 2025-03-19T18:45:04.656Z
Duration: 1 h 15 min
Description: 3 Strategies to Elevate Your Travel and Expense Management with AI

Transcript for "3 Strategies to Elevate Your Travel and Expense Management with AI": Helping you streamline process enhance, processes, enhance compliance, and gain deeper financial visibility. We'll go ahead and wait a couple more seconds just to let some people join, and then I'll go over some housekeeping items for today. Great. Joining me here is Jenna Jolley and Marcus Souser. I will be sure to introduce them here to you all shortly as well. But before we jump into our webinar, we do have a couple of housekeeping items. So first up, Imbursed in Motion is back. We have received quite a bit of interest and space is limited, so be sure to register today. If you're interested, you can find more information under the docs tab. Our next webinar will be on April 17 around ERP and expense management integration. We are calling this a master class, so we have some really great information, and hope you can join us. You can find more about these events under the docs tab. Speaking of, you'll find the docs tab and the engagement area on your right. We also included some additional resources, some good blog posts, and content for you guys around AI that we think you'll find relevant to date today's topic. At Imbursed, we encourage you to ask questions throughout our webinar. You can do so under the q and a tab at any time. Chat is also available today, so feel free to engage with us and your fellow webinar attendees. You as a reminder, this webinar is recorded as well. You will receive a copy of the webinar recording twenty four hours after today's event. To kick us off here, we always like to open our webinars with a couple of poll questions. So I'm gonna go ahead and open up our first poll question for you guys. To participate, the there is no submit button. All you have to do is select the quest or the answer that is most relevant to you. And so our first question here is how would you describe your organization's current approach to AI in expense and travel management? Do you find that you are more risk adverse and stable, conscious but committed, fast moving and impact driven, or bold and strategic? And it looks like as the questions come in here, we'll give you guys a couple seconds to answer. But we are seeing some that are risk adverse, maybe a little cautious. That's definitely the norm when it comes to AI, and we'll be talking more about, that in our conversation today as well as why that might be the case and how you can overcome that within your organization. Perfect. We'll give you guys a couple more seconds here. Definitely lean in more towards that cautious but committed category, and we'll go ahead and close this poll and move on to our next poll question. Our next poll question, is what is your organization's biggest challenge when it comes to adopting AI for travel and expense management? This is going to be multiple choice, so feel free to select any answers you feel relevant, and I'll open that up for us and then share those results here. Great. As a reminder, it is multiple choice. Feel free to select multiple options. Looks like we're seeing quite a few are saying that they have concern about data security and compliance. Good news for you, we are also going to be talking about that in our webinar today. Seeing some, also saying difficulty integrating AI, and lack of trust. We'll be talking about all of these, so definitely stay tuned for, some some ways to address those. And we'll give you about a couple more seconds before I close this one as well. Perfect. And that poll has closed. Thank you for your participation. In today's webinar, I'd like to introduce you guys to our speakers. Marcus is our senior vice president of product management at Imburse. Marcus has been in the expense space for fifteen plus years now and heads up the AI and data analytics efforts at Imbursed from a product perspective. He works hand in hand with engineering to apply AI solutions across the organization that drive real customer impact. And in addition to Marcus, we also have Jenna, who is our product marketing manager here at Imburse. Jenna works closely with product experts and customer teams to connect common pain points in travel and expense management with value added solutions. One of her focuses is on launching our new AI powered capabilities that help customers solve for what's next. And with that, I'll pass this off to Marcus and Jenna to tell you about the importance and influence AI has within your travel and expense management programs. Alright. Awesome. Thanks, Taylor. I'm really excited to be here with all of you today to talk about how AI is changing the game for finance teams in ways that traditional automation really only scratches the surface of. AI, while many of us are still maybe risk averse in embracing it, it's it's no longer a futuristic concept. It is here. It's practical, and it's changing how we work. And today, we're gonna explore how AI can help finance and travel teams eliminate inefficiencies, improve compliance, and make smarter spending decisions. But before we dive into AI driven strategies, let's just level set. What exactly is AI and why should finance teams care? So at its core, AI or artificial intelligence is the ability for machines to perform tasks that traditionally required human intelligence, whether that be humans doing the work or putting the rules, in place, to help automate workflows. So this includes understanding language, identifying patterns, predicting outcomes, and handling process with little to no human involvement needed. So in the world of travel and expense, AI streamlines expense processing by automating receipt capture, enforcing policy compliance in real time, detecting fraudulent activity, and providing predictive insights to optimize spending, all with minimal manual effort. But what makes AI technology different today, and why why is now the time to adopt it? So as shown here in some findings from Software Oasis, we're seeing a massive jump in AI adoption by finance teams in the last three years with 85% adoption expected this year. AI, like I said, is no longer futuristic, but really essential for finance teams for several reasons. So for one, AI is finally mature enough for even finance teams to trust. It's no longer experimental. AI tools are now accurate, scalable, and built with enterprise grade security. We're also in a time where finance and travel teams must do more with less. So economic uncertainty, tighter budgets, hiring slowdowns mean that teams need to be more efficient. There's also growing regulatory and compliance pressures. AI can help companies stay compliant by automating audit trails, flagging anomalies, and ensuring reporting accuracy, and to my previous point, doing so in an efficient way. Additionally, the office of CFO is becoming more strategic. The role of finance is shifting from bookkeeping to business strategy. AI is able to enable, real time financial insights, allowing finance teams to act as strategic advisors rather than number number crunchers. And so with all of this in mind, AI adoption really is no longer optional. Organizations that leverage AI in travel and expense management gain a strategic edge, and businesses must either adapt or fall behind. It's no longer a nice to have. It's a necessity to stay competitive, compliant, and efficient in today's fast moving business environment. So what are some areas ripe with opportunity for AI to elevate your travel and expense management? So to do more with less, infuse intelligent automation in receipt and invoice processing to eliminate busy work. In the face of mounting regulatory and compliance pressures, enforce policy compliance in real time, reducing fraud and manual audits, and to become a strategic business partner, leverage predictive spend insights to improve budgeting and cost control. So over the next thirty minutes, we'll break down each strategy a bit further and provide some actionable next steps and considerations. So I'll pass it over to Marcus to talk a bit about the evolution of receipt and invoice processing. Thanks Jenna and welcome everybody. Let's start out with the first component, which is really one of the obvious ones to talk about. The extraction from data from any kind of document, receipts, invoices, and so forth. And it's gone through many iterations since the beginning, and it it kind of makes sense this to be front and center because it's sort of a foundational, piece, when it comes to AI or or any kind of expense or investment management system because as the data is extracted, the as the accuracy goes up, really everybody in the life cycle downstream benefits all the way up to analytics. Right? So if you go back, to yesterday, I really drew up sort of a timeline here. It was a lot about manual data entry. Right? The user had to manually enter, information about the expenses. And then over time, there was the introduction of, something called OCR, which really just means getting turning an image into raw information. And then alongside of that, there were simple rules implemented back in the days to say, hey, if it says total in large letters on a receipt, the number next to it is probably the total amount. Right? That's how it started. And then it evolved gradually over time until today, where it really has gone way beyond just OCR. Right? The AIs that we're using really today in those systems are very much subject matter experts and know everything around the tax laws and the different taxes in any kind of country globally, the different dialects of the languages in the countries, the local merchants in those countries, and even simple things like date masks and currencies, of course. Right? So they understand the context of what they're extracting and can even infer some of the missing data pieces that aren't even on the receipt itself. So what is the result of that? Much higher accuracy, of course. Right? You can roll it out globally, even if you're a US business, let's say, it's a huge benefit to have this available globally because you might travel anywhere. Handwriting has improved significantly, the extraction of handwriting. And also more and more fields can be extracted much faster now. The bottleneck is not at the extraction, but rather in other places today. Right? So the extraction has really, stepped up their game, if you will. The benefits of that, of course, more automation and workflows, more efficiency, eliminating, errors of manual entry and coding and so forth, saving managers time, finance team time of rework and review time, essentially. Right? Okay. So that's where we are today. Now then again gradually, but now in my opinion faster, we'll go over to the next stage. Right? It's not gonna take ten years anymore to get over to the next stage. And the first thing you'll see is adjacent capabilities to be up leveled. So let's say the merge between receipts and cards, that will go way up in accuracy. Normalizations will go up way in accuracy. All the sort of adjacent capabilities will now, step up their game. And then you will see very likely in a not too distant future, systems autonomously grab additional data points and triangulate them. So as an example, if, engines today can already infer that, okay, this meal has probably a certain number of guests based on the entrees that it can find on the meal. Right? So then the system might go out, look for other expenses within your organization to triangulate, okay, who else was maybe at that dinner and figure out to fill in the gaps. So at the very least point out to the user, there might be some attendees missing here or too many and so forth. Right? So those things will happen or it might reach out to existing calendar integrations and so forth to pull more, threads in terms of data. But before we get carried away, let's go over to to Jenna and talk a little bit about actionable, takeaways. Alright. Yeah. So, in thinking about really where you can start taking action today to get the most out of AI driven automation, the first step is to step back and assess your current process and identify the biggest bottlenecks. Are expenses getting stuck for long periods in certain approval steps? Are employees still entering receipt data by hand? Where are the inconsistencies arising? Is your team spending time chasing down missing invoices? So by taking a close look at these inefficiencies, you can start to pinpoint where AI driven automation can replace human intervention and significantly speed up approvals and payments. And once you've identified areas for improvement, the next step is to, implement AI powered receipt and invoice capture. Unlike traditional OCR or optical character recognition, which Marcus was mentioning, which simply extracts, text, AI driven OCR is highly accurate and can understand context and infer meaning, code expenses, and flag missing or inaccurate information. And this means employees no longer need to manually input standard fields like date, amount, merchant, location, or category, or correct errors. So AI does the heavy lifting, getting it right the first time and eliminating tedious data entry. And then finally, AI is most effective when it operates in a fully integrated ecosystem. So by connecting AI powered OCR with corporate card feeds, automated receipt imports, travel booking tools, and more, organizations can ensure that transactions auto merge, receipts are captured in real time, and reconciliation happens seamlessly. And this not only reduces errors and provides a more complete picture of spend, but also eliminates the need for manual intervention in the review and approval process. And this integrated ecosystem can continue to grow and expand over time as we get to a place where there are more and more of those AI agents Marcus was mentioning where you can talk between systems and and pull in more of the the broader picture around, spending activity. So, stay tuned for for more there in the future, but there are things you can do today to start to connect, all of that data into kind of through a single system. So while fully autonomous expense management is still on the horizon, these steps ensure that your organization is AI ready for the next phase of innovation. So now let's move on to our second strategic strategic focus area, enforcing compliance and reducing fraud. Right. Yeah. So also a very important component, of course, is making sure the spend is within policy and without any fraud. Right? So let's see how innovation, move the needle here. So if you, again, go back to my my little timeline there. Organizations yesterday really translated their policies into hard coded rules, configurations, and went through a sort of semi manual, enforcement process. Really, there was an auditing team or maybe in smaller organizations, just managers, but often in your organization, auditing teams that then looked at all of the expenses or a portion of the expenses, to review them, against fraud, for example. And to an extent, that is still very much the case. Right? These hardcoded rules still have very much have a place today. Gradually, though, that improved in in '2 different areas, from yesterday to today. So, one of the areas is it just became much more convenient to put these rules in place. Right? There's QA or UAT systems that you can now put these rules, you can play with them. You have version control often. You can then promote them to production, revert them again if something didn't go the way you had planned. That's one thing. The other one is especially post submit. There's a lot of audit AI driven audit capabilities now that are being put in place to augment the audit team and the workflow. Right? So as a result, much better tools, better checks, and so forth, fewer human errors because of that, and the benefit is obviously smaller auditing teams, cost savings because of it, or higher level of compliance, or both really. Right? A side product is also, via these, sort of integrated auditing, panic recognition, and so forth. You get now important insights to maybe go back, in the front end of the process to adjust some of your policies and rules. Right? So that's today. However, still very much a pre, presubmit. You have hard coded rules to make sure you align with policy, and then post submit is your audit area. Right? So that is still very much this two step process placed today. So tomorrow, as we, gradually migrate over to tomorrow, we believe that more so than not, policies, risk assessments, fraud prevention will be autonomously understood by the system much more and applied wherever it makes sense. So not necessarily this split between pre and post, but much more gradually applying, sort of rules or audit checks where they make sense. So as give you, to give you an example that might make sense, and and these things are rolled out sort of as we speak today, is a check around, itemized meal receipts, for example. Right? So if you're at the restaurant, you're capturing a meal receipt, you might wanna have a notification as a user to say, hey, make sure to capture the itemized receipt. Letting the user know right there and then where it makes most sense, where it's most efficient to get that notification. Right? Or in Europe, for example, if a VAT receipt isn't attached right there at CAPTCHA, or as a user submits their expense report and puts in their business purpose or applies to a policy, warning or violation, to make sure that the response is sufficient. AI is very good at understanding the context and making sure that there's sufficient information so you reduce that that back and forth. Right? So these are just some examples, of how we think that is much more dynamic in the future, and that also helps, help build confidence in the system as we then go to a world where we more and more see automated creation and submission of expenses happening in the future. That's a good positive side effect. So let's look at some of the actionables from from Jenna. Alright. Yep. Thanks, Marcus. So in thinking about how your organization can take advantage of today's AI capabilities to enforce compliance in real time and prevent fraud before it happens. The first step, similar to what I recommended earlier, is to do a bit of a retrospective and look at past audit reports or compliance violations to evaluate where compliance breakdowns are happening today and where AI driven monitoring can have the biggest impact. So some things to think about, are employees unintentionally violating policies because they're unclear? Are certain categories like meals or mileage more prone to out of policy spending? Or maybe is there a suspiciously large volume of spending funneling through the miscellaneous category? We sometimes hear that. So by analyze analyzing past productivity or common, policy risk kind of violations, finance teams can identify high risk areas and refine their policies to ensure they're both effective and easy to follow and even consider if the timing of the compliance checks is in the right place in the process, like Marcus was mentioning around perhaps, being able to notify user earlier in the process so that in real time, they're able to ensure that they're in compliance with that policy. And then AI helps by continuously learning from historical trends and flagging recurring risk patterns, ensuring the organizations enforce the right policies at the right time as well rather than applying rigid, one size fits all rules. And, additionally, a major shift AI enables is proactive compliance enforcement. So instead of catching non compliant expenses after submission, AI can flag policy violations as employees enter their expenses like we are saying. And this means that employees are notified of out of policy spend as soon as it's captured, whether it's an unapproved vendor, over budget, missing a receipt. Maybe it's a duplicate transaction. So it allows them to correct those issues before submission. So this not only helps prevent fraud, and reduce policy violations, but it also makes compliance more intuitive and seamless for employees and really helps to minimize that back and forth between submitters, improvers, and finance teams, that nobody's happy about at the end of the day. And then finally, AI is transforming the audit process. So allowing finance teams to focus on high risk transactions rather than having to review every single expense manually. I AI powered auditing tools like our partner, AppZen, analyze spending patterns and flag anomalies, automatically. So that ensures finance teams not only, only review truly suspicious activity rather than low risk transactions. Inverse audit as well further enhances this by combining some of those AI insights with human expertise. So with AI, finance teams can automate compliance monitoring, reduce manual workload, and catch fraudulent activity before it leads to financial losses. So AI tech AI isn't just about, making compliance more efficient. It's really changing the entire approach to risk management by preventing issues before they happen. So now let's move on to our third strategy, how AI can take financial insights from passive reporting to proactive cost control. Alright. So the third strategy, of course, we have to talk a little bit about what you do with all the data. Right? So the analytics side of things, How did that evolve over the years? So really back in the days, finance relied heavily on historical data, for making budget decisions. For example, doing manually analysis of identifying certain trends, cost saving opportunities. For example, travel managers manually compared flight costs to spot overspending. Right? So essentially, systems provided the raw data for analysts, reports, and simple dashboards, to do that that research. And that's still very much in place today. There's still very much a spot for that and it's even required today to some extent. Right? So I don't think that's gonna go away. One of the most common, reports or reminders really that we still see is the unsubmitted corporate cards, for example. Right? So these things are still very common. It's not to say that these go away. Now today, softwares have much more targeted insights that offer takeaways, really, where then a finance team or administrators can go double click, find out more, take action on these insights, and improve the process, really. So for example, there might be vendor analysis dashboards or reports that analyze overspending trends, right, in combination potentially with benchmarking. So one can really look at renegotiations with the vendor or look for alternative vendors, right, reducing in cost savings. Or what we also see already in part implemented today gets to sort of the point underneath, going from more passive to proactive cost control, reshopping tools that automatically, monitor when flights or or hotel bookings drop in price and then automatically reshop them at a lower price. Right? Obviously, reducing in in cost savings without much manual intervention. And then as we shift to tomorrow, we're gonna go more and more, I'd say aggressively in towards, from a passive to a more proactive cost control concept, that allows finance to act even even faster. Again, everything before is still very much important and it's still very much in play. But what we foresee really in the future is, for the analytic systems to really package up recommendations in a much more concrete, sort of actionable way, alongside maybe some KPIs and metrics as well. So you really can think about it as, okay, we identified, some insights here that you might take action on. Here are some KPIs tied to that and some configuration changes to go with it. So really, you get this recommendation and you can implement it right away with a click of a few buttons, and you can measure it over time, maybe for thirty days. You can roll it back again and so forth, similar to what we talked about in the previous section. So to give you some real examples, simple ones. Right? So maybe the system realizes 98% of rideshare related expenses meet the approval requirements. So maybe you wanna do a simple auto approving, put a simple auto approving rule in your workflow. Right? Very simple example. Now it might come with some KPIs that are measured. Maybe we're measuring the the auditing, behaviors, at the same time, and we are proposing this configuration change that you can just execute and roll in and see, how that, improves things over the next thirty days. Right? Or we the system might notice high out of pocket spend in certain groups of the organization. So the system might automatically recommend for these individuals or groups, you might want to issue some virtual cards or issue corporate cards, right, and package that all up for you to take action on. Again, just some examples here. But, yeah, let's see what Jenna thinks on actionable takeaways on this front. Alright. Yep. So to fully harness AI powered spend analytics, finance and travel teams even can start with really three key steps, identifying top strategic questions, optimizing data for and with AI, and integrating AI driven insights into workflows. So AI can surface a ton of data. And if you think about the nature of travel and expense management, it's there is just the volume of data that is being captured, in in those types of systems is tremendous. But so to make it truly actionable and to ensure you're tracking the right data to begin with, it's important to start by defining the right questions. So what is it that you need to know to make better financial decisions? Where are you overspending? What policies are driving compliance issues? What cost saving opportunities are being missed? The key to is to align AI driven analytics with your most critical business priorities, ensuring the insights you get are both decision ready and strategically relevant. For instance, a university may ask, how are we effectively managing grant funded travel to ensure compliance with funding regulations? Or a manufacturing company may want to know how does RT and E spend vary across different production sites? And are there any cost efficiencies in key regions? A law firm may be interested in what percentage of travel is reimbursable by by clients, and how can we improve tracking and cost recovery? For professional services, firms may wonder, are certain clients driving disproportionately high T and E costs compared to the revenue they generate? And then finally, for a growing corporation, they may want to investigate how can we standardize travel policies across newly integrated business units. So, obviously, a lot of different areas you can focus on taking your data, and it can become analysis paralysis. If you don't have alignment as an organization around based off of your initiatives and where you're looking to head in the next few years, what are the key things we really need our data to be answering? So moving on to our second takeaway here, better data in, better data insights out. So if your financial data is inconsistent, unstructured, or missing key details, AI driven analytics may struggle a little bit more to deliver reliable insights. So organizations should really ensure their expense categories, policy rules, and vendor data are structured and standardized in their expense management tool, really aligning with system best practices. But this is a two way street. So, the good news is that AI is also going to help to enhance data accuracy and visibility. It can infer and be able to interpret meaning, take and take context to fill in some gaps, can improve categorization, can normalize fields like vendors or categories, and help to refine financial forecasts over time. So the better your data, the more precise and actionable your AI driven insights will be. And then the final step is ensuring that AI powered recommendations don't just sit in reports, that they actually lead to real business action, especially as we're able to do more and more with, some of those proactive AI recommendations and insights over time. So it's important to to establish, a process where AI generated insights are regularly reviewed, escalated, and converted into cost saving actions. So this could mean automatically flagging spend optimization opportunities, recommending vendor renegotiations, or maybe proactively alerting teams to budget risks before they escalate. On the end user side, this could mean ensuring employees receive AI powered spend recommendations when planning trips, such as selecting, preferred hotels based on corporate rates or booking cheaper alternative flights. So AI driven insights really should be embedded into everyday decision making, ensuring that finance teams can act on them at the right time and in the right context. So now that we've explored how AI can drive automation, compliance, and strategic cost control, let's go ahead and wrap up with some final thoughts on overcoming those barriers to AI adoption that, I know a lot of organizations on this call have indicated, may may be in place. Alright. Great. So, as you begin some of these strategies, you might come across, common barriers or common concerns that slow down adoption and and, you know, in part might be valid concerns as well. Right? So there might be resistance to change, securities concerns, you brought that up during the poll, that was a big concern. Justification for investment, if there's additional cost associated with it. And these are all normal challenges that that you might encounter. Right? They don't have to necessarily be roadblocks. So I wanna break them all down a little bit, see how we might wanna overcome them, and walk you through that a little bit. So on the the first barrier, really resistance to change is often about being unfamiliar with the technology, difficult word for me, being unfamiliar with the technology, fear of maybe job displacement, skepticism about the accuracy, Right? So without fully understanding it, it's sometimes difficult to see the the, the necessity of something that might appear more complicated to begin with, rather than a tool that that really helps add value. So one of the things that, especially as as we started with our new team around AI, and I tried to push the team for is really be hands on. Play with the technology. Back in the days, a couple of years ago when ChatGPT was announced, I tried to push everybody to use it every day. Just play with it, push your workflows towards that. So being hands on, training employees, that's a good first step. Then also highlighting the benefits. How it can eliminate manual data entry or reduce approval bottlenecks. Also, focus on maybe phased rollouts rather than a full scale rollout all at once. Participate in pilot and early adopter programs. There's usually in organizations such as ours, lots of early adopter or pilot programs you can sign up on and you can be part of building it out rather than being afraid of it, for example. Right? I think that is also a good step. And, one important one is also the continued dialogue around it. Right? So if you have people that are afraid or worried and have concerns, resisting, continuously talking about it and having a dialogue is very important. Right? If you have somebody that goes to a conference once a year about it, that is probably not gonna help. Right? And I assume you also have you're confronted with AI and the benefits of it across many different software applications you implement, not just expense or invoice management. Right? So it's something generically that you probably wanna think about. So the second barrier, the important one you mentioned that earlier in your polls, of course, is we're processing sensitive financial data, personal data, and so forth. So, privacy and being compliant, is obviously a big concern. Right? That's a need, for an organization that they want assurance of. Right? And so this is very important and very valid. Right? So you always want to make sure the organization that you, use software from has robust encryption standards, has a strict data governance policy, right, is compliant with all the different industry standards, GDPR, SOC two, and so forth. And especially when it comes around AI, also if they're using third parties, making sure there's zero data retention policies with that third party. There's no uptraining based on your data with a third party. And ideally, there's at least a path to, building out, own models on the infrastructure of the software that you purchase. That is usually the route the industry is going right now, so that is a good step to observe as well with whoever you purchase the software from. And often, companies also have, things like a trust center, right? So a separate place, a specific place online that you can, after you log in that you can enter maybe, that has all the information about data privacy, security, and so forth, and usually also an expert you can talk to. Alright. The third barrier is, the finance team is sometimes hesitant. Right? When approvals or fraud detection or forecasting is done automated, that's often the question of, well, what is the accuracy really like? Right? There's maybe a lack of visibility around the decisions that are being made. So it's important to just find out, explainability or transparency about these systems. Right? Why do they act or how do they act? And maybe also ask for some transparency on some of the quality metrics. Right? What does the accuracy look like for me and my system specifically? Right? Is there anything that is being tracked? Maybe include human oversight in the beginning, right, which you always be have when it comes to data extraction from receipts or invoices in the beginning. The user will always look at that still potentially or auditors. Right? And the audit tools that are in place today, the auditing team usually is in the loop as well. Including human oversight in some of this transition is is very helpful as well. And lastly, leadership often wants to know, okay, what is the ROI on this? Right? What's the benefit or the value behind, these AI solutions? And if it's not clear, then often these things are deprioritized. Right? So again, seek solutions with transparency around this. Often they have analytics platforms tucked into these solutions that might be able to visualize some of the cost savings. Right? And some of these things are still continuously evolving, especially right now as this innovation is really kicking into gear. Some of these companies are still trying to figure it out. They're still adding metrics, associated with AI powered, automation. And so keep an eye on that. Even if it's not there in some of the solutions you have today, it might come very shortly. I know it's a priority for everybody, I would say. So keep that in mind as well. Right? It might not be all there day one. Keep an eye on it. Again, join early adopter programs. So that's kind of my, for all of the above, sort of my advice or recommendation. Be hands on with these things. Stay on top of the developments that are happening. Be open minded. Maybe incorporate human oversight. Take a closer look. See if there's any metrics out there that are being provided. And if you have serious concerns, I would assume in all of these solutions that you might use today, there's product teams that would love to talk to you, get your feedback. Right? So be involved. They don't have all the answers necessary either, and they want and need your feedback. Right? So open up that dialogue, and yes. So let's, let's summarize a little bit of what we talked about today, and then maybe also do a little bit of a future outlook on some of the user experience changes that might happen. So we talked about some of the new and exciting AI related capabilities just to take advantage of right now to drive efficiency, enforce compliance, gain deeper financial insights. We also highlighted some of the concrete actions on how to get the best out of it for your organization by automating some of the receipt invoice processing, strengthen policy enforcement, leveraging predictive spend analytics, and how you can move maybe from a reactive financial management to a more proactive strategic approach. Right? So we talked about all of these things. And now I wanna leave you maybe with a little bit of an outlook of what the user experience itself might look like in the future, how that might change. Right? We talked a lot about these other components, but what does the user experience look like? And so especially as you talk about a very AI empowered software and how the interaction with the user works. So going back sort of of my timeline that I used before, if you think about yesterday, you an AI, you might think about chatbots or assistance. You might think about Siri and Alexa. Right? I always like to use that as example. If you talk to Siri and Alexa today, it seems integrated. Right? But if you'd ever and that's why I kind of wanna leave you with some homework maybe. If you've ever used ChatGPT in voice mode, you know what I'm talking about. If not, that'll be your homework. Try that out. Right? Maybe you have a ChatGPT subscription, maybe not. If you have the subscription, you have even this advanced voice mode. And if you click on it in the bottom right corner, I think it is, it pops up this circular bluish bubble, and you know you have advanced voice mode. And it has a very fast reaction time now, and you can really have complex conversations about any topic. Doesn't have to be expense management. It can be any of your hobbies. Right? You can talk about finance or macroeconomics, and you can have a very complex conversation, and it understands really everything. And so I'd really encourage you to do that to see what it what the AI can do today to kind of get a glimpse into the future. And so going back and why am I talking about this? Right? So going back into the expense world, there's hundreds of different use cases that that we wanna perfect for the user and improve upon. Right? Whether that's per diem in Germany or certain tax data that needs to be entered or mileage that needs to be entered, attendees that need to be added. Right? There's hundreds of different use cases. And even if all the other things we talked about earlier are done through 99% accuracy, there's often still certain, certain information you need an answer to, right, as a system, as a software to finish the expense. So for example, I live very close to the airport here in LA, and my wife sometimes drops me off at the airport. The system just doesn't know that. So at the very least, the system might reach out and say, hey. Did your wife drop you off at the airport again? Yes. It yes. She did. Okay. Perfect. I'm I know your address. I'll put that in the system so I took care of that for you. Right? So at the very least, you'd wanna have that interaction. Or somebody joins, Jenna joins a dinner I have with my team last minute. The system can't know that she joins last minute. Right? So I might have to add the attendee manually. Right? So for all of those interactions, there's also a lot of the UX that will change in that respect. And will it be voice or chat? Well, we don't know necessarily. Right? That's still out there in the open. Many people like it, some people don't. That is still questionable. But walking through a couple of instances, if you're a traveler, for example, and let's say we wanna ask that mileage question or example I talked about, I might be in in meetings, I might be in Zoom, I might be in email right now or Slack or Teams, and the system might reach out to you and say, hey. Can you quickly confirm that you actually had your wife drop you off at the airport? Or can you confirm this booking that I wanna confirm for this event? Yes. Yes. Go ahead. Okay. Great. So that might be what an experience looks like. Right? Or as an approver, you now might not be in the loop at all anymore, whether you're a manager or an auditor. But the system might automatically, pull you in and ask you via Slack or Teams to say, hey. Can you maybe just confirm that that John was actually was he supposed to fly business class? Was that okay? Yeah. That was okay. Don't worry about it. So those things might happen. Or as an analyst, right, rather than going into analytics and pulling up a dashboard, he or she might just send a message into Slack or Teams against the analyst, bot, and and and that might just generate an image right from scratch there. Rather than a dashboard, you get an image for a graph automatically generated from scratch from the analytics platform. Right? So those are just some examples of what it could look like. But just one thing I wanna leave you with and encourage you is, lean it. Right? The change is probably gonna happen. So participate. Be an early adopter. Be hands on. Ask questions. And, yes. So I hope you enjoyed the sessions, took some away from it, and, yeah, keep the conversation going. And let's look at some questions here. Yes. We got a lot of good questions coming in. Before I pull up one or two here, while we have the time, I just wanna point out that we will follow-up if your question doesn't get answered. And if anybody is interested, we definitely can combine all our responses for the larger group with some later outreach here. But let's make sure to at least address one or two of these questions because we got some really good ones here. So I'm gonna pull up, this question here, from Suraj. Does your AI feature use existing data to improve performance automatically, or, do you expect customers to work with you in sharing this data? How do you guys typically see that approach, with customers looking to implement AI technology? Yes. So it's a great question. So some of these models what I talked about earlier, especially when it comes to data extraction, they know everything then is to know about receipts and invoices. So it's really not the same anymore as it used to be that you have to uptrain these machine learning models. The engines really know a lot about these documents already. So it is not that you really have to, you know, confirm as a user or correct, a certain amount for the engine to get better over time. That is really not how it works anymore. So it's a good question, but the mechanisms are different than they used to be. So we you can also opt out from utilizing data to up train some of our internal models, but it's also not needed to the same extent anymore than it used to be, if that is helpful. Great. Thank you, Marcus. Yeah. We we also got quite a few questions, around you had mentioned flagging noncompliance early. Is this different than, a compliance rule? And can you elaborate a little bit more about that, what you mean when you say flagging noncompliance? Right. So, really, we don't wanna make away with the rules that that exist today. Right? There's always a place for a hard stop. Right? There's no more than like, expenses above this amount should require receipt, for example. There's, you know, there's no need to replace that in any sense, if that makes sense. But there's things that we can put in place that are more on an automated fashion. For example, we can, for some of our law firms, for example, we could say, hey, we extract the alcohol from the receipt and we can automatically deduct that from billable amount. Or we can, flag if you have, you know, if we believe that on this year, they are actually there have been fewer attendees that you added to these attendees. We might flag that. Right? Because you might wanna reduce the kind of average spent a little bit and trick the system. We might flag those type of things automatically. Right? So things that you don't really put into the system manually today that we might wanna capture. Now we're always gonna give you the option to opt out of some of those things as they introduced, but these are a couple of examples, related to that. Perfect. Thank you, Marcus. And, yes, we are at time today, guys. So, if we didn't get to your question, we will be sure to follow-up with you. Please be sure to also check out our April webinar, around ERP and expense management, and, we thank you all for joining. Thank you, Marcus and Jenna, for such an insightful presentation. We hope to see, all of our attendees at our April webinar. Thanks, all. Thank you, everybody. Thank you.