Real Talk: What to Expect When Implementing AI for Customer Support

Watch our webinar where our founder Sami Ghoche and Head of CX Rose Wang, go over seamless AI implementation for Customer Support team, how AI enhances you customer’s experience and much more!

Want to learn more about Forethought? 

Webinar Transcript

Megan Murphy: (03:48)
All right, good afternoon, everyone. Hope everyone’s having a wonderful day. Happy Tuesday. A fun fact, it’s National Puppy Day. So if you have a puppy, thinking about getting a puppy, regardless, I hope maybe you could leave this webinar, Googling them, but glad that you all are choosing to spend your afternoon, maybe morning and maybe evening with us, here at Forethought. And so my name is Meg Murphy. I’m the director of sales here at Forethought. I’m here today with two incredible people, Sami Ghoche and Rose Wang, both Sami and Rose are Forbes 30 Under 30 recipients, and also graduates from Harvard University. Sami is both our co-founder here at Forethought and our CTO and Rose is our incredible head of customer experience. Together they bring a breadth of knowledge and a deep understanding of AI from a technical perspective, as well as a digital transformation perspective.

Megan Murphy: (04:45)
And so we’re all really excited to get into the weeds with you today. We’ve got a lot to run through and cover. And so whether you’re just dialing in, trying to learn about AI for the first time, you’re in the middle of exploring AI for your company, or you’ve recently implemented AI and you’re trying to make sure that you’ve actually done the right thing for your business, our goal today is to leave you with a few things surrounding AI, which is one, understanding the real definition of AI, two, the impact that AI has across our customer experience, three, how does AI impact support agents and four, where do you begin when it comes to implementing AI? And so without further ado, I just want to jump right on into it, unless Sami and Rose, you have any fun facts to share with everyone dialing in today about yourself, that that they may not know, based off of your incredible profiles.

Sami Ghoche: (05:41)
I’m not a very fun person, but I think one fun thing is, I used to do competitive Rubik’s Cube solving. So that’s kind of fun. I stopped.

Rose Wang: (05:56)
That’s really cool. This is so funny. I fiddled for a year. I grew up in Nashville.

Megan Murphy: (06:05)
What? I can’t believe it. I had no idea, actually. I hope to pull it up next time we all can meet together, but cool. So Sami, let’s start with you. What is the actual definition of AI?

Sami Ghoche: (06:23)
Awesome. Thank you, Megan. First of all, just want to say it’s a pleasure to be speaking to all of you today, going to keep it very understandable. None of the big buzzwords that actually don’t mean very much, hopefully. So just simple definition of AI, big buzz word these days. As opposed to more traditional software that you can think of, AI is a system that does not have to be explicitly programmed to return specific outputs to some given input. Instead, behaves more in a smart way, using quotes deliberately, just more the way that humans actually do and perform tasks that otherwise normally are thought of as requiring human intelligence. So pretty basic definition. And it’s probably also what most of you have in mind when you think of AI. I just wanted to plant a flag in the ground there.

Megan Murphy: (07:21)
Okay. Yeah, no. Cool. I love that. And thanks for again, setting the stage here on how we define AI here at Forethought and what the actual definition is. And it seems pretty straightforward, but how does this actually fit into the customer experience and across the businesses today?

Rose Wang: (07:39)
So when AI fulfills its promise, everyone wins. I’m going to paint you a picture of a before state. So everyone has dealt with customer experience and customer support before. You have a ticket come in. And it is a ticket where a customer filled in an arbitrary category. That ticket gets routed to the wrong queue. And then it takes hours if not days, to get a response to the customer. And then what’s the after state? Well, the after state with AI is that we can answer your tickets in seconds, especially the questions with answers, somewhere in a knowledge article or a macro. Now with automation workflows, we can even perform simple actions like refunds and cancellations for you, faster responses that are accurate because you get to control the accuracy, which we get to talk about later, results in a much higher customer satisfaction.

Megan Murphy: (08:31)
Got it. So it seems like tweaking the workflows to basically remove the mundane. And again, making sure that the answer is that we are spitting back out, from an AI perspective, are accurate and would seem as if it is a human that’s providing that experience for you instead of an agent itself. Okay, cool. Actually, that’s funny that you bring that up. The other day, we were speaking with a company at Forethought who was looking and exploring AI across their customer support team for the first time. And when we asked them what their goal was, we hear a bunch of different companies share their goals with us across the sales floor here at Forethought. And it’s really interesting to hear about what are their objectives, what are their OTRs for their business? But the one that really stuck out to me was that, in going back to your painting a picture, was their ultimate goal is to never have a support ticket, hit their queue ever again.

Megan Murphy: (09:24)
And what that meant to them as a business was that, it would have meant that they would have solved the problem of a seamless customer experience, where you could live in a world and engage with a company and be self-sufficient without having to wait or be delayed and interacting to get the answer to any question that you have or any troubleshooting that you needed to do. And so I thought that was pretty incredible, the way that they explained it and articulated it. Because that seems pretty outlandish to think of, but with AI and the way that the world that we’re living in today and the digital transformation that’s going on right underneath our nose, I don’t think we’re that far off, from being able to interact with some companies where the self-service experience is actually really incredible. So now going towards thinking about the way that the support agent role will change, how does AI impact your agents, Rose?

Rose Wang: (10:17)
Think about the life of an agent. They are getting the same questions over and over again, constantly going from tab to tab, to find what they’re looking for, after reading negatively-toned messages over and over again. It’s tough.

Sami Ghoche: (10:30)
And actually the average annual churn rate for agents in the U.S., agents in U.S. context centers ranges between 30 to 45%. It’s double the average of all other occupations in the U.S. Obviously a big problem. You invest all this time, months of effort into onboarding and training new agents, let alone recruiting them. You’re investing this amount of work and money, basically to have them become experts. Once they finally hit their stride and they’re performing, they’re doing great. That’s why you hired them. They might only stick around for three to six months before they eventually leave. Also critically, when they do leave, all the expertise that they’ve acquired and accumulated over all that time, is just gone with them, which is doubly bad. And that’s except if you use Agatha, which is something that we’ll cover in a bit as well, but definitely a big problem.

Rose Wang: (11:30)
Because what Agatha can do is, it can help by taking care of the simple, repetitive tasks, refunds, cancellations. So agents can focus on more complex cases that require human judgment and empathy. Plus our Agatha Assist product helps onboard your new agents and speeds up the process of researching and answer and crafting a good response. We’re a complete platform that follows your tickets from the customer to your agents. That’s providing much more context than a one-point solution.

Megan Murphy: (12:00)
I love hearing that, thinking about an agent, especially in today’s remote world and environment. Most people I speak to, if they’re starting a new job, they really want to feel tied to the mission. And so when you feel like you’re doing the same repetitive tasks over and over again, the idea of having that be removed from an even state a day, and that they’re actually rolling up their sleeves and really working on some of the challenges across the business. Or maybe as you mentioned, providing an empathetic response to a customer that’s important to a company, ties that employee, regardless of them being there for one week or being there for 10 years, 20 years, to the mission, having higher retention of employees, overall employee satisfaction and whatnot.

Megan Murphy: (12:43)
And just again, the way that the world is developing and has changed and been forced to change, I love the fact in thinking about supporting as being more empowered, to really impact a company’s bottom line. So cool. So where do we start here? So it sounds like we’ve got different ways that support agents are going through digital transformation. There’s different workflows that are going to be tweaked by leveraging AI. There’s a lot of things that go into making sure AI, it works well when implemented, not just from Day One, but as it gets the maintenance that it also requires too. So where do we even start? Let’s begin there.

Sami Ghoche: (13:19)
So I mentioned AI being this big buzzword. AI for customer service is also, you can think of it as a pretty noisy space. All these chat bots out there that you’ve probably heard of, and this has been going on for quite a few years, but I think things are really changing in just the way that AI fundamentally can be used more effectively. I like to classify the companies out there that mentioned AI for customer service, into three buckets, broadly speaking. So the first bucket, there’s companies that claim the smart behavior. Again, we said AI qualifies as AI, if it’s doing things that are thought of as critical thinking, things that humans typically have been able to do. But in fact, what they’re actually doing in this first bucket is they deploy bots that require a support manager or admin to actually define a ton of mappings of basically inputs and outputs manually, which is what we talked about as being opposite of AI.

Sami Ghoche: (14:20)
So as a result, managers will have to dedicate a ton of time, to perform, add these mappings, know which mappings are required in the first place, actually use the software, new software. They have to get acclimated to put in the mappings. And inevitably, they’ll miss lots of cases, which ends up meaning that the AI solution is only as efficient as the person who ended up configuring it, which is what you want. You want it to be helping the support manager. You don’t want it to be lagging behind them and only getting whatever they remember or get to give the AI solution, in their spare time, so to speak. So that is the first bucket. Definitely can lead to negative customer experience, as you can imagine. This is the scenario where some keywords can really lead to very bad answers that you can actually understand how quote unquote dumb, but it can be performing that way.

Rose Wang: (15:30)
The second bucket of companies, they actually will deploy bots that can be qualified as being smart and can be called AI. However, the AI that they use is actually closer to proficiency at dealing with general language. So since general language is still one step removed from the customer support world and two steps removed from the specific customer support issues that you face at your company, these bots won’t be able to tackle the nuanced questions and leverage all the experience you’ve accumulated over the years of serving your customers.

Sami Ghoche: (16:03)
We’re in this third bucket, which is the bucket where our AI is actually trained on your history of thousands, or even up to millions of support tickets that you might have. Tickets, chat transcripts, the questions that you get asked, the way you’ve answered them for years, those data are a treasure that you’re probably not harnessing properly today. I mentioned previously that when your best performing agents leave, so does the knowledge that they’ve gained, but it actually does still sit there in the CRM, it’s just, no one can actually access it or access it in an actionable way.

Sami Ghoche: (16:37)
Agatha can ingest this years’-worth of that expertise, except she can actually do it in a few hours. So all the experience she needs is actually already there in your CRM and your tickets and your external data sources that we connect you with, ready to be put to good use. So that’s actually how we can call ourselves experts from Day One. That’s why our philosophy is to rely on your support history instead of placing on you, the burden of teaching our AI how to act properly. We do the heavy lifting, Rose is going to talk a lot about that, but we do have another thing. We still want to make sure to give you visibility into how our AI operates and make sure you can control its behavior.

Megan Murphy: (17:20)
Interesting. Okay. I like the way that this is broken up into three core buckets, essentially. It’s moreso of talking about, what is the extreme. I wouldn’t want to go against Agatha in head-to-head competition as a support agent, to be honest from Day One, but I appreciate the extreme of being able to read all this data in just a few hours and then be able to take action on it. And then also the idea of not having to, from an implementation standpoint, as purchasing AI solution for my company, thinking about implementation and then how do I actually maintain that when my business changed, how do I have to direct retrain these models? Do I have to make sure that I’m tweaking the type of language that my AI is understanding? So let’s talk about what implementation with AI looks like then. If you want to start with that, Rose, we’d love to hear not only your perspective of here at Forethought, but overall in general, what you’ve seen implementing AI at a new organization.

Rose Wang: (18:22)
Absolutely. It’s what everyone’s here for. What does the implementation actually look like? So I’ll go into it. Chett, can we get to the next slide? Awesome. So what we’re really proud of here at Forethought, is what Sami talked about. Our models are trained on your data specifically rather than relying on general language, which has many steps removed in our problem you’re trying to solve. So what does this look like in time? Well, the first step we do is that we, once we go through security review, we ingest your data. We want to index about 20,000 of your past tickets, or if there’s more complexity to your tickets, we want to index hundreds of thousands of tickets. So our models then have enough training data to actually output something very accurate. And so basically what Agatha is doing in this moment is, she’s learning much like humans do.

Rose Wang: (19:10)
This only takes about two hours, much faster than humans, at which point, our team switches to discovery mode. And so this is really where digging deep into the CSM that you have for your company, the implementation engineer that you have on the team, really important. So our team goes really deep. And in this discovery mode, we dig into your data to identify areas where we can help reduce manual tedious tasks. One of those could be solving questions with macros on email or knowledge articles on a web widget. Another could be a new prioritization of your queue with tags that Agatha applies herself with much more efficiency and accuracy based on how you’ve applied tags in the past. And this takes our team about one to two days to complete. Then once we’ve identified potential areas to automate, we’ll then check in, with your team, before we run our models on your data.

Rose Wang: (20:00)
So for example, if we want to predict your email spam tickets and close those out versus having agents open, label, and then close those tickets, then all we need to do is look for past spam tickets, train our models to look for incoming tickets that are similar in content. Then we’ll identify spam tickets and then we’ll close those out. That takes about one to two weeks. Then the model goes into testing for a few hours and when you’re ready, it just takes about 30 minutes with you on Zoom to get the solution live. After that, we’re just on call to monitor. That monitoring stage is really important. We don’t just implement and walk away. We want to make sure, Hey, as tickets come in, are we making the accurate predictions? If not, do we need to go and tweak? And we catch that within really, we’re looking at that hours, not days, we’re not letting that sit. And then in total, that’s it. That’s about one to two weeks, which makes us the fastest AI solution in a support industry to deploy.

Megan Murphy: (20:59)
Love that. And so it also sounds like there’s a lot of interaction between your team, Rose, and then also, you’ve got Agatha of the AI solution that is across all Forethoughts products, and then you’ve also have your team that’s dedicated to an account, but it seems like there’s a lot of human interaction as well between what the business cares about, and then also what they want to see Agatha be able to do. One, how do I trust your team, but also how do I also trust Agatha as a solution?

Sami Ghoche: (21:35)
A great question. Definitely something people ask us a lot. It’s nice in theory that we do all this heavy lifting, but just this idea of a black box that’s hard to interpret that’s doing something really smart under the hood, how that would be daunting. How do I actually know what you’re doing and know that it’s good.

Sami Ghoche: (21:53)
So I like to talk about three things. One of the very nice consequences of the fact that we do train these models on your own data, is that we can actually perform what’s called replay analysis on your historical data. That means after we’ve trained our models, we can run them silently in the background on new incoming tickets and give you basically spot-on estimates of the performance you should expect, if you were to deploy them. We can tell you here’s the accuracy that we expect. Here’s the percentage of issues that we’re going to be tackling or automating, or depending on the use case. You don’t have to go into this blindly and just deploy our solution and hope for the best. You will have predictive analytics before we go live. That’s the first piece and is very much related to the way our AI works.

Sami Ghoche: (22:44)
The second piece is that you can actually test our solution behind a white list. So if there are certain agents or admins or managers who should be testing the workflows that we’re implementing before, obviously, you roll it out to all your customers or even all of your agents. That’s something that we offer. So deployed things behind the white list could be email addresses, could be based on IP addresses, firewall, you name it. And just double check that it’s providing your customers and agents with the experience that you want and that you can be proud of.

Sami Ghoche: (23:24)
And then if things are good as they normally are, you deploy, but you feel confident in doing so. And then the third one is, I think there’s an important distinction to make. We’ve been talking a lot about how we do the heavy lifting, which is great. Machines can maximize your workflows in ways that typically, your team of experts can’t on their own because we’re looking at the whole thing available 24/7, but we do offer many avenues for you to control many aspects of our solution. So we offer an admin dashboard.

Sami Ghoche: (24:03)
You can get predictive analytics broken down by knowledge, article, template, response, or workflow for example. You can decide everything from which piece of content to turn on or off or for each of our products, what filters to apply based on, for example, account or customer tier information. And you can control how confident you require us to be in order for us to automate anything. And you can even do this at a granular level. So in an interpretable way, you can actually configure our AI to intervene or not intervene, based on really the use case. And then the type of experience that you want your agents and your customers to have.

Megan Murphy: (24:51)
Love that. Thanks for all the details, Sami. One of the things I keep coming back to is, you’re talking through the way that you train the models, the way that we think about data, it sounds like you have to rely a lot on data points across the company. And one of the things that we hear quite a bit is that, let us get the data perfect. Let us get our process perfect. But what happens when a company’s data actually isn’t ideal? What then?

Sami Ghoche: (25:20)
Quick question. In practice, the world is never perfect and we’ve had so many customers, that’s typically the case across the board. Some are maybe more organized or more mature than others, but it’s my job, it’s our job to make sure we can still provide a great solution even in that situation. So a lot of what we’ve invested in, is actually being able to have our models detect the signal, even when it’s being drowned out by noise. Without going into boring you with the details, but you can imagine if there’s a specific way, things are being answered by some experts agents, but a lot of the time, it’s being answered a different way, that answer is spiking up as far as our models are concerned. And that is the answer that our models will still like, even though it might be drowned out by a ton of inconsistent answers or labels or whatever the case may be.

Sami Ghoche: (26:20)
So that’s number one. Good AI can actually still tease out the signal. Number two is that we have some advanced machine learning tricks that can actually help us learn from smaller amounts of data. So an interesting example is that we can actually take things that might be written in English, let’s say, translate them into another language, then translate them back. That’s actually called back translation. And there’s some differences between what we started off with and where we ended up. And that actually can be thought of as more data that we can train on. So that’s just like a flavor of the types of things that we can do. And lastly, I’ll hand this over to Rose, but our implementation engineers are experienced. This is what they do. They’ve done this hundreds of times. They know what to look for. And we have this discovery phase after which they’re coordinating with your team to double check some of our assumptions. So we’re making sure to take into account any knowledge you have about how your team currently is set up and where we can trust the data and where we maybe shouldn’t.

Rose Wang: (27:35)
Thank you, Sami. Yeah. Well, I think what’s so incredible is once we get a customer, they don’t typically go elsewhere and it’s not just because of our AI, but it’s because of the team that we have with implementation and our CSMs. Really, your solution is as good as your team is willing to dig into the workflows and know your business. So let’s take an example. So oftentimes, let’s talk about email, right? We want to respond with a macro to an email. What our team will go do is, we’ll actually read through all the different tickets to see, Hey, when a question is asked, what did we respond? Once we go through about a hundred, 200 of those tickets, we have a really good sense that, Hey, did you know that 20% of your customers, the reason why this macro didn’t answer their question is because they had another question that you should just add into your macro and it will help cover another 20% of your tickets.

Rose Wang: (28:35)
And that way we’re increasing your deflection volume, but we dig so deeply into your information that we know even better, the customers themselves, about what issues they’re getting and how to tackle those issues.

Megan Murphy: (28:49)
Yeah. It sounds like having that out, that third-party perspective. A lot of times, I think companies, it’s like when you get a new hire and leadership, or even maybe a frontline employee, and they give feedback on a process or the way to think and do something differently, than what’s in place today. It sounds like you all encourage your team to do that, learn about the business, but really ask why it is the way that it is, and be able to tweak it and make it better while leveraging the technology of Agatha.

Rose Wang: (29:25)
Absolutely.

Megan Murphy: (29:26)
And so for anyone that’s actually dialed in, if you have any questions, feel free to chat them in, on the side, into the chat window. We’d love to also run through any ones that are coming up, that are top of mind that maybe you have not been with, to ask maybe to someone that you’re speaking to, or as you’re doing all of your Google searches as well. Sami and Rose, as I mentioned, bring a wide variety of knowledge here, both on a digital transformation perspective, but also from the AI and tech stack perspective. So feel free to ask away. I’m going to ask this to you, Rose, what has been one of the most impactful things that you’ve been able to do for business since starting here at Forethought? Looking at all the customers that Forethought has, these are some incredible logos. And so talk to me a little bit about some of the changes and ways that you’ve been able to help companies leverage AI.

Rose Wang: (30:23)
Love that question. One thing I do want to point out is when you look at our logos, that is something I’m so proud of. One, you can see that there’s not one industry we service, that there’s not one size of company we service, but really as long as you have enough data, which for us is about 2000 tickets a month, we can help you. And so that’s something that’s so beautiful about AI is that we’re really industry agnostic. We’re size agnostic, workflow agnostic. We’re here to help. And so I can tell a story of, from big to small, it really depends from, even just, let’s say you’re a customer that gets about about four or five different issues that come in, whether that’s cancellation, refunds, et cetera, most companies still have agents going through and answering those tickets.

Rose Wang: (31:17)
And they see a lot of churn in their workforce because it’s a very repetitive task. It’s quite boring. And so there’s no reason your agent should be answering those questions. It’s not fun for them. And then that also then affects the customer experience. And so with our web widget, what we can go do is we can pull from your macros, your knowledge articles, and answer those tickets about refunds, cancellations that are already there. We can even go and plug-in to the backend of where you hit the refund. We can connect to that backend and hit that refund ourselves and perform that action. So for those companies, we’re seeing at least about 20% deflection volume, where 20% of the tickets that come in, or that customers ask questions for, they can find it through Agatha because we help them find the answer in their knowledge base or macro.

Rose Wang: (32:07)
Another really exciting one that I like to talk about is, we have a customer who is one of the top 10 data storage companies in the world, Fortune 500 company publicly-traded, and they have an amazing C-SAT score for being a publicly-traded company. Part of the reason why is, they have a very great process, a manual process today where any customer, any sales rep, any support agent can escalate a case. And that escalation, what it does is it triggers a process that is extremely costly, upwards of four to $10,000 per escalation.

Rose Wang: (32:47)
And if you get hundreds a month, which this company does, that’s millions and millions of dollars, not only in direct costs, but also indirect costs, to go and make those customers happy. And so what we’re able to do now for them, Sami is the one who built this, is that we’re able to go in and identify and predict a case that’s going to escalate before it gets escalated. And so then we basically prioritize their queue and it’s not so simple. It’s not one thing that causes an escalation. It’s 20, 30, 40 things and all weighted differently. And so our model is able to take very, very complex decision-making, escalate cases and save this business about 4 million annually.

Megan Murphy: (33:35)
Crazy. And yeah, when you think about that from at least my perspective, I’m very much task-at-hand, think about it, having multiple data points, some that you don’t consider. And the fact that Agatha can leverage the incredible AI, to make better decisions for our company and have that type of impact, is truly amazing. My question actually is going to go to Sami. Obviously starting in the AI world and looking at the market, there are things like chatbots, there are things obviously like Agatha and there’s all different definitions of AI. What is the thing that you’re most proud about that Agatha has been able to do or what you’ve been able to accomplish since creating Agatha?

Sami Ghoche: (34:23)
I’m extremely proud of the team that we’ve built, first and foremost, including our implementation team, which our customers deal with day in, day out. I’m very proud of the technology that we built. I’m biased, but that’s one of the big things that I wake up and I’m happy about. I think we started off, we’ve learned a ton, but we started off just focused on one thing. Initially it was, let’s augment agents. We mentioned all the issues that they face with onboarding and performing repetitive tasks, not being confident enough about the answers they might be giving. So our initial product was actually just Agatha Assist, which is about assisting and augmenting agents. Over time, our customers actually pulled us into being this holistic solution that helps out across the support staff. So that’s where we started adding our second product, Triage, which is a product that routes incoming issues to the right team or the right person, sets the priority level like Rose mentioned, potentially preemptively escalates things.

Sami Ghoche: (35:42)
After that, customers also pulled us into Solve, which is our third and most recent product where we automate the fall case resolution cycle without a human in the loop. That, we don’t do that for everything. We do it for these repetitive tasks that Rose mentioned. And so we actually, today, provide the entire platform and we’re rapidly building things out and getting better and better, but that’s one of our key advantages. Often the companies that we’ve talked to, love to consolidate their vendors. They don’t want an AI for this, AI for that, bots for this. And that’s something that today I can say we offer where we didn’t before. I think that’s something that sets us apart.

Megan Murphy: (36:34)
I love that. Yeah. There’s a lot of technology out there. And so having everything on one and thinking about going back to the beginning of this, the idea, I always think about never having a support ticket hit your queue again, and providing that self-service model of, every single customer should be able to find the answers that they’re looking for, if leveraging AI correctly and servicing of that information in a timely and a very clean way, but that’s also clear for them to understand and can take action on. So any other questions or any questions from the attendees that are here? I know that there might be some, but I’m definitely here to answer any other questions that you all have and make sure that you guys are getting everything that you need. Rose, is any final thoughts that you’d want to leave everyone with or same with you, Sami?

Rose Wang: (37:40)
What I love most about Forethought and this team, is how much each of us care, truly. I think we’re a little obsessive with our customers in a sense that we take their business on as if it were our own. And I think it doesn’t really matter who you go with, but looking for that level of care in the vendors that you work with, is so important. And that’s something that to me is not just a job, but this is my purpose. And so it’s really important that we can bring this solution to you to make your life better. That’s really what I see us doing.

Sami Ghoche: (38:16)
I can’t follow that up any better. So would love to hear any questions from the audience if there are any, but otherwise I’m excited to potentially be working with some of you in the future. Any additional questions, please go to www.forethought.ai And please ask for a demo. We love talking to new customers.

Megan Murphy: (38:45)
Love that. Thank you, Rose and Sami, for sharing your knowledge with every single person here and also appreciate the technology that, Sami, you’ve built from the ground up and alongside of your team, but also to echo what Rose is saying. It’s really incredible to be able to sell a product that, from a technical standpoint, checks off all of the boxes as it pertains to AI and machine learning and the capabilities, but also that the human element and the empathetic approach that we are able to provide our customers, is something that’s also really important, especially in today’s world. And you can have a digital transformation, but if you lose that human element, what is the actual purpose here? And so, Rose and her entire team are dedicated to all of our customers and future customers. And so thank you for attending today and spending your afternoon. Again, happy National Puppy Day. And I hope you all make everyone else’s day wonderful. And thanks for choosing to spend it with us here today.

Sami Ghoche: (39:45)
Thank you.

Rose Wang: (39:47)
Thank you, everyone.

Think forward with Forethought.

We’re here to make enterprise-level AI second nature for your business.

View demo
Call to action decor Call to action decor