Hila Segal and Roni Kandel on Learning Arc, AI Training, and Learning in the Flow of Work

Russ Fordyce 🏆 (00:01.522) Welcome back to the winner circle everybody. I'm here today with Hila and Ronnie from Walkme. They are changing the way that we are training our teams and training our staffs and they are the latest award winner. So Hila and Ronnie, welcome to the winner circle and congratulations on the award. Hila Segal (00:18.467) Thank Roni Kandel (00:19.244) Thank you so much. Great to be here. Russ Fordyce 🏆 (00:21.844) So before we get into all the guts of this, you guys have developed this kind of new product that really frankly I could have used throughout my career because I was complaining about training all of the time. people would come up to me and say, you know, why doesn't sales understand this? We train them. And you guys are solving for that, which I really love. So Hila, why don't you just give us kind of that elevator pitch on Learning Arc and Walk Me. Hila Segal (00:45.827) Yes. So again, great to be here. Thanks for having us. I'll start with Walkme. Walkme really pioneered a category called digital adoption platform. So we're really over the course of the last, I say, 13 years have been helping employees in the enterprise actually use the software, navigating workflows, driving adoption of scale of all the technology that organizations are bringing into their business. About 1,600 customers globally, about 27 % of them from the Fortune 500. So for years, we've helped these employees really make sure that they're navigating this technology, these workflows, they're doing it in the right way. Helping them really with the execution of these workflows, day in and day out. But our customers kept telling us that They had a deeper need to build more skills and lasting proficiency right in the flow of work, not just in a separate sort of training session or even just a portal that may be disconnected from where the work actually happens. And that really is what brought us into sort of the learning space and really what the, you know, that's really what WalkMe Learning Arc was designed to solve. So WalkMe Learning Arc is really an AI native learning solution. really sort of built from the ground up on AI, and Rony can kind of talk more about this, it combines very sophisticated content creation that is all based on AI with the ability to deliver that learning content inside the application where people work every day and to really reinforce the learning. at the exact moment of need when people actually need this. And this can be done on any device in multiple formats and across any workflow that you're working at. So you may be in sales and you may be in a different department and you may have different training needs. And we're able to really use that sort of understanding that Walkme has developed over the years of what's happening on the screen. What's the context? What are you trying to achieve? Hila Segal (03:08.895) and be able to service and surface to you the relevant learning content. So in a way, we're sort of catering for the authors and for the LNDs of the world who are really tasked to make sure that they can help their people keep up with the pace of change and make sure that they're building the right skills. And also on the other hand, making sure that our learners, right, they're getting that learning content that they need right when they need it. Russ Fordyce 🏆 (03:36.654) Yeah, this is such a huge problem. I spent my career kind of training salespeople, training teams, customer service organizations about product, about even the process flow of how things work, the tools that we use. And in a traditional environment or in the olden days, we would go into a conference room, you'd have a PowerPoint, you'd show them the tool, and then they would go off and try to do their job. And maybe that thing would come up. But generally speaking, it wouldn't. It would be... a week later, six weeks later when they needed that information. You guys have really kind of changed this and flipped it on its head. Kind of more moving away, I think you guys say, knowing to doing. So walk me through a little bit about kind of how you guys got into that concept of doing versus knowing. Hila Segal (04:26.943) I think it's a combination, right? So as I started Walkme for years, really helped people do the work, do the work in the most, in the right way, less errors and more efficient way. We're now sort of complementing this, so you're able to actually get some of the, are we doing this the way we're doing this? So one great example for us is actually from the world of sales, right? So I think you've... you've brought up this example. So we all roll out sales methodologies, right? Every company sort of has their own sort of preference. And that sales methodology is basically asking our sellers to do certain things in a certain way. The way that you may be, you know, submitting your forecast or the way that you may be, you know, progressing an opportunity from one stage to another. So Wacme can help the seller right in the moment when they're actually changing and updating an opportunity, changing status, moving it from one stage to another, Walkme can be there to help you make sure that you're doing this in the right way. But at the same time, Walkme can also surface to you some relevant learning content, a module that really explains to you why is it done this way and what kind of information we would like to see from you. about the competitors, about the decision makers, about all those other important parts of the sales methodology. So that's how we're combining the doing, but also with making sure that we're explaining to you how and why things are done this way. Russ Fordyce 🏆 (06:12.365) So you guys have built this, as you mentioned, you had this WalkMe product, which I've used before. I'm sure a lot of our audience has where a screen comes over your screen and it gives you tips on what's going on on that screen. You guys decided to build this learning arc as a separate product, not just a feature of WalkMe. WalkMe through that decision and why you guys decided to do it as a standalone product. Roni Kandel (06:39.608) So I think our main goal, like you last said, is bridging that gap between the learning part and the doing part. And we still see the need. When we talk to our customers, we see the need to train sometimes before you give access to sensitive systems or sensitive processes. And that actually completed that entire package and helped us to bridge that gap. We allow them to train beforehand. And then when they go into the system, It's first off, it's not the first time they've seen it. They've been trained before. They have a bit of a familiarity, but they're not expected to remember what they've learned. That learning content resurfaces. And I think it gives a lot of peace of mind that it's not the first time you're encountering this. It's not the first time you're trying to do this live. You're getting that same experience that you got beforehand while you're actually doing the work. Russ Fordyce 🏆 (07:32.652) Why do you think this has been so persistent? We've had this problem for such a long time of kind of training and then actually execution. Why do you think it's taken so long for somebody like Walkme to work at solving this problem? Hila Segal (07:51.203) You want to take this, Loni? Roni Kandel (07:53.198) That's a great question. Why did it take so long to solve this problem? I think the two fields have been completely separated for a very, very long time. You had domain experts, you had the L &D experts, they were focused on training employees to do their job. And then you have those kind of IT software managers, they were in charge of deploying the actual application. And there was no easy way to connect the gap between those two. I think we're inventing a new era or a new set of skills and responsibilities and empowering L &D leaders to take bigger ownership and bigger control on how learning is actually being done and also to see the results of that learning. We're aiming towards removing those kind of vanity metrics of just usage and if they've done it, that means that they know the material and we can directly link the learning to actual achievement within the work that's being done, within the applications, within the processes. So we're linking all of these dots that were separated before and can now prove that learning can be efficient when it's delivered in the right way. And you see that direct link between the learning and how the work is being performed. Russ Fordyce 🏆 (09:19.795) Yeah, it's all about the context, of when and where you need it. Go ahead, Hila. Hila Segal (09:25.326) I was just going to build on this a little bit and just talk about some of the changes that I think is what's happening around us right now, and especially in the workplace, right? The pace of change has really outgrown some of the traditional sort of learning programs, right? And because there's so much that is happening all the time, you need to constantly rescale. You need to constantly build. you know, more proficiencies across your, know, across your sales team or your overall sort of workforce. And we're seeing that the content often goes to sale. You know, as Ellen B is trying to sort of, you know, create that content very quickly. We're seeing that there is not enough reinforcement and the ability to really remind people and sort of give them more opportunities to really consume that content in the right places. you know, so these are the conditions of the market are changing so quickly that there is more demand, right? That it's coming at what learning programs are designed to do. And I think a lot of this is the kind of problems, right? That we have set out to solve with that product. So on the one end, we're... you know, allowing and sort of empowering more scaled content creation with the power of AI that allows the teams to create content, you know, in a faster pace, maybe refresh and update existing content as well, and make sure that it's constantly aligned with the business objectives and sort of what the company is trying to achieve. And on the other end, trying to, you we often refer to this as re-imagine with digital learning experiences are for the learners and, you know, create those experiences that are more integrated into their daily work and, you know, allow them to consume this learning inside of their, inside of their applications and maybe in multiple formats. So really Hila Segal (11:43.94) This product is catered and designed for both of these audiences. Russ Fordyce 🏆 (11:48.941) Yeah, I remember, you know, unless you were taking me somewhere fun to train me like Florida or some, you know, warm destination, it training is always a chore. And, you know, as a, as a product marketer, know, you always had to do it. It was always on your list, on your checklist to do. And, you know, once you gave that training, you kind of check the box. I trained sales, but that was never, you know, that was just the beginning of the training journey. And so when you started to roll this out to kind of early customers, early beta customers, how did you know it was working? Hila Segal (12:20.74) So far, we've seen this product launched late last year. So we've been steadily growing the customer base here. And what's exciting is the breadth of the use cases that we're already seeing among these customers. have Customers that as Lonnie said before are really going through major transformations and want to make sure that they're training the teams ahead of these transformations. A lot of times this involves large volumes of content creation and really sort of pushing it to, you know, big teams and large, you know, parts of their organizations. So we're seeing, you know, teams using this to really eliminate some of the L &D backlog and create more content quickly and deliver this in app. We're seeing teams that are doing this to de-risk some of their major ERP rollouts, for example, and provide more structured onboarding to drive faster adoption of these new processes and new systems. The common thread that we're seeing is these are organizations that have scaled needs for their learning purposes. And they're really trying to find ways that can accelerate and create more efficiency in the way that they drive their learning programs. Russ Fordyce 🏆 (13:54.028) Well, and you're dealing with an audience, the LND teams usually, that are weary of new solutions sometimes. And you've really got to build trust in my experience with those teams, either from the content side or from the tool side. How are you finding that the LND teams are kind of responding to learning arc? Roni Kandel (14:14.99) So I think you're spot on. We saw a lot of hesitation in regards to AI usage when we started off and started to communicate with customers and L &D leaders. And first, I'll say they were blown away by the results. And to chime off a bit about what from what he last said, the first comment is always, this would have taken me months. And we've now done it in a single session in a few minutes. So that always kind of wins that first trial. But then the next question is, OK, now let's see the accuracy. Let's see the content. And I have to say, think that's where we really nailed it and we bought their trust because not only are the results truly impressive, we give them full visibility on what the AI did. In any element that they're seeing, they can dive in and see exactly what the AI did, what was used from the assets that they provided and how closely linked it is grounded to what they've provided this entire AI. That gave them peace of mind and a lot of peace of mind and lot of trust within the system and a sense of control and ownership, which is extremely important when you're in charge of such a process. Russ Fordyce 🏆 (15:14.57) Yeah, it's almost... Go ahead, sorry. Russ Fordyce 🏆 (15:30.676) Well, it's really genius because I'm just thinking about it. You're building a little custom agent, almost a little L &D person inside of that AI that is kind of living with the person, watching what they do and trying to understand how to better serve them. And I like what you guys said about being in the flow of work. What does that actually mean inside of Learning Arc? How does that represent itself? Roni Kandel (15:57.944) So it's actually one of the easiest flows that we have. When you create the content, you can manually decide where you want it to be surfaced, which section of it you want it to be surfaced. And as we'll evolve and move forward, we will integrate more and more behavior-based AI-related ways to surface that content. So as Hila said, we already have the knowledge of when a user is in a specific process, areas of process which are prone to errors. And then we will continue to automatically surface content based on those known areas that are prone to mistakes. Hila Segal (16:34.545) So I'll just build on that Russ, because it's Walkme, we have a lot of information about what's happening on the screen. And this is based on a technology that Walkme has been developing over years, this deep understanding of the screen as a human would. That level of understanding allows us to really understand the context and then be able to... respond to you and surface to you different learning content that maybe is relevant for the moment. What we can also do is, and this is actually an example of something that we rolled out here inside of WalkMe, which is, everyone's rolling AI tools and we wanted to, we have created an AI literacy training course inside of WalkMe Learning Arc. And we wanted to make sure that everyone is, you know, before they jump in to use some of these AI tools is, you know, has gone through the course. So here we can, before you're about to enter and maybe open your cloud or Gemini or any tool that you're using, we're able to, you know, remind you and nudge you or all the way up to just, you know, blocking you, right? And telling you, Hey, you haven't actually. Roni Kandel (17:43.95) you Hila Segal (17:56.926) completed that course. So there is obviously the help that is provided to the user and making sure that they're really getting that relevant training. But also from the responsibility of how we want our team members to be using these AI tools, that level of literacy is important. So there is opportunities for us to do to really surface that learning content. And in addition, one of the things that we've done is, know, Ronnie and her team spent a lot of time thinking about, again, we talked about how do we reimagine this world of digital learning, and even in the portal. So part of the capability of our product is that we do have a learning portal where you can log in and you can find learning content and you have learning that is assigned to you and you can really... track your progress, you can find any new types of learning content that you're looking for. That portal has been designed in such a modern and new way that was really intended to create a new experience for the learner. That you're not coming here just to see a folder that will list some courses. for you, but it's a real experience that is inspired from the type of experiences that we all have in our private lives, how we consume our media content. So Ronnie, I don't know if you want to talk a little bit about that, because I thought that was really inspiring in the way that we've thought about this and really putting the learner at the center. Roni Kandel (19:50.572) Yeah, completely. I think L and D leaders have always had this kind of Sophie's choice, right? They had to choose between either make the content fast or make it engaging for the learners, right? If you want to make it engaging and fun, it takes hours of work and you put everything you got into it. In this new AI era, that blocker is out the window. There's no reason to make that choice anymore. It can be easy and fast. and super engaging and fun and amazingly flexible. So we're just leveraging off of that and enjoying those capabilities. The fact that we're brand new, we're starting from scratch and maintaining that flexibility that no matter what you put into the system, whether it's a PDF, a video, an image, whatever you'd like, the output can be flexible as well. The output can be an audio, a video, a text, and the user can now choose. So it doesn't feel like a chore anymore. They can choose if they prefer to consume it through an audio, through a video, through a text area. Because we all learn differently. We have different preferences. Russ Fordyce 🏆 (21:00.622) Well, you're preaching to the choir on this. It's amazing to me. With me and my children, you look at them and look at how I learn and look at how they learn. We all learn differently. I'm a very visual learner. I'd rather watch a video, have somebody talk to me than read. My oldest son would rather consume everything in written format. And that's the way he digests it. You guys did a kind of practical thing. And I think a lot of AI companies are doing this. to where you've got this AI element, but you've also kept humans in the loop, as they say. Walk us through that decision and how are humans still involved in this other than just deploying the content. Roni Kandel (21:44.014) I think that was one of the easiest choices we had to make. We're targeting large organization. The L &D oversight is probably the most important thing at this point in time. In the same way that IT want to control the types of softwares you have in the organization, it's the L &D's domain ownership and expertise that will define if learning within the organization will be effective. So at no point there was a question on maybe we can do this completely automatically, no supervision is needed. It's just the opposite. We want to extend the reach and empower those people with the knowledge on how learning should get done and help them focus on the core principles of how learning should get done and minimize the effort that they need to spend on things which are, you know. designing it, making it more engaging, adding flashcards and videos and stuff like that. We want them to focus on core principles and the AI can do the rest. Hila Segal (22:51.97) I'll just, I was just going to add to that, that, you know, we've, we've, we've really thought carefully about how AI applies into this craft, of, of, creating learning content. And one of the areas that we've really, invested in is the control and the precision that anyone that creates content can apply here. Roni Kandel (22:51.979) Not bad. Russ Fordyce 🏆 (22:52.364) Now there's a, go ahead, Hila. Hila Segal (23:21.5) So this is not fire and forget. You put a prompt, the AI will create it for you. There's a lot of mechanisms that were built into the product at different layers and levels of creating a course, all the way from the outline of the course to the individual components of the course. The simulation, the quizzes, everything that sort of comp- comprises a learning module. And we've provided a lot of tools that allow the authors of this content to really fine tune and make sure that there's no hallucinations and the AI is following my exact instructions. In some cases, I don't even want the AI to use. the general purpose LLMs and I want to be very sort of prescriptive to what I'm asking, you what are the sources we're allowing the AI to, we're pointing the AI to very specific content. You know, a lot of times we're not creating learning content from scratch, not sort of a blank page. There's existing content. There's, you know, processes and PDFs and videos and things that have already created. And going back to what we talked about, which is the pace of change, you need to refresh this. Maybe there's new procedures, some new things that are coming into the org. Doesn't mean that you have to start from scratch. So this has been designed when I said at the beginning, this was not, you know, an existing learning product that we bolted on some AI capabilities. This was really built from the ground up on AI, but Roni and the team has spent a lot of time thinking about how do we create that experience? How do we take the best of what AI has to offer? But we're contextualizing it and we're bringing this to the world of digital learning and creating learning content. And that makes that product pretty unique in the way that we have thought about these things. Because I think it's a very natural question. You asked this before about some of the hesitation and kind of what was the reaction of some of the Hila Segal (25:44.692) learning professionals. This is probably the most common question that we get, right, Loni? Like, how am going to know that the AI is not going to produce some bad quality content for me? So this was very intentional to make sure that this content is not only great on the first run, but there is this ability to continue to really fine tune it to make sure you're getting the best outcome. Russ Fordyce 🏆 (26:19.486) There's always this, I think you're bringing up a really great point, which is there's this tension between kind of that personalization and scalability, or there has been, right? That usually in my world, the more personal you got, the worse it got, right? And as you scale that, it just got even worse, right? And a good marketing example is, you know, if you want to look a company up, you can look a company up and understand it does manufacturing, right? They have a little code next to their, you know, Dunn's number or whatever. that tells you they're in manufacturing. But as you drill down on that number to see what kind of manufacturing they do, it gets worse and worse and less accurate and less accurate the finer you go. How did you guys resolve that tension between personalization and scalability? Roni Kandel (27:06.776) First off, have to say, I think we're constantly improving that. I think it's something that will work with us as we evolve constantly and we'll continue to evolve and fine tune those mechanisms. The easiest solution is always that human in the loop that does some of the oversight. And on top of that, we provide flexibility, I think, to the learners where if they don't feel comfortable with one then they could just switch to another. So if you start off on a video and it doesn't feel right to you, or it's too granular, you can go back and go to, let's say, the text mode if you're like your son and you prefer to read it. So you don't have to choose. We provide you with the flexibility. I think that sets the expectation where they should be at this point in time and maybe makes the decision a little bit easier. Eli, if you have anything to add on top of that. Hila Segal (28:06.053) Yeah, I would just say that in addition to the content itself, think an element or dimension of personalization is also what we talked about before, which is the delivery in the moment that this is most relevant to me. So that's also an element of that. And, you know, that's kind of the approach that we're taking. And as I've mentioned before, there's a lot of information that we're able to leverage. in terms of segmentation, knowing who you are, understanding where are you and your workflow. And that allows us to provide that level of personalization as well. Russ Fordyce 🏆 (28:53.311) If we zoom out a little bit, you guys are enterprise software developers, which I love. You're my people. If you think about kind of the larger impact that, that the tool that you've got to build this tool and it's, it's, it's just mere existence tells us a little bit about kind of where the enterprise software world is headed. Give us a little bit of vision on, outside of learning art kind of where companies are headed with their development cycles. Hila Segal (29:26.739) That is zooming out for us. listen, I think that we're going to start seeing and already starting to see rapid, rapid innovation. There's just a lot more power today in the hands of developers that allows us to experiment, to get to market a lot faster than before. And that actually applies to the entire life cycle of software development. You know, from the ideation to the way the product is built to the way that it's wrapped and, you know, the design elements of it to how we market it. So, you know, we're going to see and continue to see a lot more faster iterations and more software that is being released to the market. Maybe not all of it is going to be great. But definitely what is not great, you can find out very quickly and just pivot and introduce new products. that's sort of from where I'm sitting, I think we're going to see more and more. Roni Kandel (30:45.23) I think we're, I completely agree. think we're going to see great velocity and everything that's happening. Things are going to evolve very quickly. You know, when you hear industry expert talking, you start hearing these declarations like you won't need developers anymore. You won't need product managers anymore. I think they're, they're semi right. You know, a lot of our roles will merge into each other. Product people will do a lot more development. Development people will do a lot more product. And at the end of the line, those that will succeed are the ones that are managing to merge that in the best possible way and deliver value, right? It will all, cause speed will be, everyone will have speed. Now the question will be about the value that you're. Russ Fordyce 🏆 (31:30.846) Yeah, I can't imagine having to train people on tools that literally change daily, you know, and that's really the problem. I mean, I see it in my own work. You know, I use a lot of different tools and there's new features, new functions that come up. You'll get an email on it, you won't pay attention and then it's there. And I really don't understand how, you know, well I do now with Learning Arc and a tool like Learning Arc. You can actually feed your feeding that AI that new information and the AI is going to do the work for you, which is just genius. If you think about five years from now about how learning arcs changing this L &D landscape, where do you think we'll be in five years? Roni Kandel (32:11.992) Well, I have, I'm extremely excited about it. And I think we are blessed with a very visionary leadership and a very trusting leadership as well that is granting us these capabilities. And they've, we've jumped on the learning wagon at the exact point in time. It's ready for disruption. It has to go through disruption and everyone are seeing it. Even we're hearing it from our customers as well. And that disruption will allow us to completely change how learning is done. Like we said, effective learning happens in the flow of work. It's personalized, it's multi-module, so you can experience it the way you like. Sorry. It's no longer like one static course, a chore that we're all doing. So I really see that direction and that area that we're going. Sorry, lost my voice there. We can do that again. Russ Fordyce 🏆 (33:08.968) That's fine. Hila, anything from you on kind of the next five years? Roni Kandel (33:12.28) Thank Hila Segal (33:14.676) You know, Russ, I think it's almost like it's really hard to say what's going to happen in five years, especially where we're at today, when things are changing here on a weekly basis. The models are evolving on a weekly basis. So I think it's just going to be a great ride for us in the software business, in the technology business. you're up for change, If you get excited and thrilled by new frontiers, then it's just going to be a great experience to see how the world is evolving. Honestly, how is it going to look for our kids? What's the workplace going to be like for them? And we're just excited to be part of this. And we're laying the foundations for what learning looks like today. But we're really coming into this with the perspective of change. This has always been the core of Walkme. This is why we exist. We exist to help organizations and to help people, mostly the people that make up these organizations, navigate change and mostly change that is brought on by technology. And in that sense, we thrive on change. So bring it on more change. know, we're excited. We're here to really help companies navigate through this. And I anticipate that we're all going to continue to feel more of that change. And, you know, we'll see the product. The walk-me learning our product continued to evolve what it looks like exactly in five years. I'm not sure, but again, I know that we're there to help people navigate through this technology change. Russ Fordyce 🏆 (35:09.573) Well, used to, you know, I used to tell people, you know, Google it and, then we tell them, now go, you know, go ask chat GPT. And I really like how you guys have flipped the model, right? It's really, no, the tool is telling you how to do it. It knows what you're doing. It knows the, you know, the process, it knows the context of what you're doing and you're really flipping it on its head. When you think about that, do you see a, do you see an angle here for customers of yours to kind of turn this outward to their customers? to help train their consumers. Roni Kandel (35:40.878) Unfortunately, they've already asked. they're already know I'm joking, of course, they are asking for it and we are considering it as something we would like to. Russ Fordyce 🏆 (35:56.428) Well, Hila and Ronnie, this has been really interesting. I'm glad you guys are in this world doing great things. As a former product owner, I can't imagine trying to do this without a tool like Learning Arc right now. It would seemingly be an impossible task. So thank you so much for joining us and congratulations again on all the hard work. Hila Segal (36:15.625) Great, thanks for having us. Roni Kandel (36:17.038) Thank you so much.