Brian Bell (00:00.718) Hey everyone, welcome back to the Ignite podcast. Today we are thrilled to have Yuvraj Singh on the mic. He is the co-founder and CEO of Arusto, a company rebuilding how organizations create, govern, and distribute learning content in the AI era. Thanks for coming on, Yuvraj. Yuvraj Singh (00:15.104) Thank you so much, Brian. Pleasure to be here. Very excited to talk to you. Brian Bell (00:19.266) Well love to start with your origin story. What's your background? Yuvraj Singh (00:22.508) Well, born and raised in a small town in India, come from very humble beginnings, was a very ambitious kid from the beginning. I was one of the first kids out of my hometown to get admitted into one of the top tier schools in the US. And that I saw that as an opportunity to start a company as well to help other kids kind of do the same. So at 17 launched my first company while I got admitted into Georgia Tech. Ran that for about five years. Took it to about a million dollars in revenue over the span of five years and eventually sold it out. When I was at Tech met my best friend and now co-founder, Cain Pudma in freshman year. We were in the same fraternity. Got excited about working with each other, worked on multiple different projects. though we took very different routes. We always kind of wanted to like, you know, come back and discuss ideas and explore things together. After Georgia Tech, I worked in consulting for a couple of years in Washington, DC, got, went to grad school for my masters, worked with McKinsey for about four years. And then I was also one of the youngest advisors to the office of the president in UAE. I think one of the coolest things which kind of got me interested into this space is when I was at McKinsey, I was essentially helping large sovereign wealth funds turn state-owned enterprises into private companies. And all these organizations really did not have any kind of learning academies or structured growth and talent development programs kind of set in. So for about two to three years, I was literally building corporate academies in different parts of the world for different companies. And that's when I discovered that this is such a fundamentally broken system. From the outside, it looks that, it's going to be very easy. People have these repositories of courses, or they subscribe to these libraries. And you just kind of assign these trainings to certain employees and certain business functions and roles. And they will go ahead with it, complete that, make a checkbox. And that is it. But actually, when you go about building something from scratch, when you have to prove that it works, Yuvraj Singh (02:24.692) and it can actually help the talent within an organization. It's actually a very difficult and complicated and expensive process. So we knew that that was a problem and I felt quite excited about it having gone through the pain myself and decided that, you know, we wanted to fix that problem. And that's what we're trying to do with Aristo. Brian Bell (02:46.478) Really fascinating. that's a good founder story too, right? you know, I love the kid from the village, first first of his you village to go to a a top tier school and start a company and it's such a great founder story. you know what's interesting is the the privatization of of state assets. What's you know, we just had a political economy episode right right before this one. well Yuvraj Singh (03:13.185) Mm. Brian Bell (03:16.034) It'll be a I don't know, six episodes before y yours airs, but what was in it for the state to do that? Is it just to increase competition, lower prices, create consumer surplus kind of thing? Yuvraj Singh (03:26.4) Yeah, I think typically in most of the most of the times when like state-owned enterprises are kind of established, it's either a function of that, it's a new country where a lot of the talent focuses primarily around the bureaucracy. So you get the best of the talent from the bureaucracy to kind of run and build these machines or it's a resource rich country, which essentially the resources are owned by the government. Right. So that's the best way for them to. kind of established these entities. The problem is as the country and the economy matures, of course, those same talent pools or the same structures of governing the company don't mature at the same point of time. Right. And as a result, you have these behemoths of organizations, right. Typically managing like hundreds of billions of dollars of assets. And sometimes like, you know, I have a lot of potential to compete with some of the world's largest private sector companies as well. just are underperforming because you don't have the right set of people in the right set of roles. They probably are not trained for that specific business function for that industry. When you compare to like other people that they are competing with in the private sector. So one of the things that we see kind of globally that as countries and these economies mature, it's always better to push some of these large enterprises away from the state. and more into the private ownership. I mean, is never, historically has been the best operator of businesses over long run, right? So I think that's kind of what triggers the transition, but it's not as easy as that, hey, we're gonna turn the switch on and then suddenly this company is going to be private, right? It takes a lot of effort into taking that ownership private and then also ensuring that do we have the right kind of systems in place that now we can attract, manage and grow the same talent. and make sure that the management also is essentially accountable for continuously accelerating the performance of that company. Brian Bell (05:34.026) Eloquently put, as I drink my first cup of coffee. so you're you're actually working in the training and development of these organizations, these new corporations, and you're seeing a recurring problem. What what what was the aha moment where you're like, okay, I can build something around this. I can actually go solve this. Because you you built a company before with with revenue. And so your founder hats already, you know, your founder, you know, water finder is already attuned. Yuvraj Singh (05:37.739) Yeah. Brian Bell (06:03.924) what was the moment you're like, okay, I see I see something repeatable in venture scale here. Yuvraj Singh (06:08.48) Yeah, think, see if you look at the entire process of how organizations think about their knowledge creation, I think that process typically starts with identifying that, okay, what is it that we need to teach to a specific individual in a business vertical in a specific role? Right. Once you determine that, and that's essentially like, what outcomes do we want for them to do their job well and for them to grow? Then you identify, okay, where does this information come from? Do I have these SOPs best practices? Do I have this knowledge within the organization? Do I need to source it from external providers? Who are those external providers? Are these universities? Are these publishers, SMEs in the outside world? Are these consultants? Once I source that information, how do I, I can't just bombard everybody with same set of information. So which piece of information is relevant for which person? How should that be structured instructionally? How do I build the foundation, the base, then kind of, you know, focus on delivering concept one, two, three, do it in a manner which is not boring, which is not mundane, right? Which means that it needs to be interactive, it needs to be engaging, it needs to constantly check whether they are retaining that information or not, whether they are actually understanding it to applying it. And if they're not understanding is, is it just that they're not understanding it or what is the reason that they're not understanding it, right? Getting all of that information while making sure that you don't accidentally give wrong information, which is not suitable for that role. And if tomorrow an audit officer comes around and they ask that, like, have you actually provided the right kind of training to your employees? Right? You have the proof that yes, this is exactly how we built the training and this is how they consumed it. And here is the entire system of records through which we can prove that this knowledge is actually applicable. And we're compliant with all sorts of like, you know, regulations that might be there. Now running their entire system is actually quite a cumbersome task because it involves multiple different stakeholders. It has multiple different teams. You're using way too many number of tools. All in all, we've seen that Yuvraj Singh (08:24.768) to produce one module, people were taking, I would say, anywhere between like six to nine months spending 10 to $20,000. And they were using, I would say, maybe eight to 15 different type of tools. And a lot of the times, the main bottlenecks were that a lot of the decisions which could have been properly codified were not really codified. People were repetitively making subjective choices, design decisions. And one of the reasons was that each of the tools were good at doing one thing, but none of the tools were really good at cascading those decisions down. So if my decision changes at the top of the funnel, how does that actually impact the last step of the creation? If the consumption is poor on the learner, how do we actually take that data and not just report it, but improve the creation process the next time around? Right. So these were some of the fundamental observations and insights that we had, which we thought that the way people have thought about designing these systems has been fundamentally not shift. Like they've not fundamentally been re-imagined from the day they started. So every solution has basically tried to fix a patch of a problem, but not really re-imagined the entire workflow. So back in 2021, 2022 is when I got the conviction that you could actually rebuild the entire workflow. One of the fundamental things which did not allow me to jump into the business at that point of time was that is it a repeatable problem? And because people, would typically build this kind of stuff, maybe let's say like, once every three years, five years or 10 years, and they're not going to refresh it. Right. But I think come 2024, that is the conviction that we were able to really build on because now there are tailwinds. which actually means that knowledge has to be captured much more frequently. It exists with lot more people rather than centralized teams. And it is going to be created in lot more modalities and almost just in time that this reimagination of the entire workflow becomes a real necessity if any organization really wants to scale up their learning content creation and training development. Brian Bell (10:49.046) Interesting. So it's a reimagining of the entire content creation, delivery, analysis, workflow in a in a training and development organization, right? and I think you you touched on a really particular problem there that I'm kind of familiar with, which is you kind of shape the the content to the tools, right? Yuvraj Singh (11:04.031) Right. Yuvraj Singh (11:14.732) You Brian Bell (11:16.114) you I'm very familiar with this at Microsoft. You know, Microsoft, we had so much training. you know, I managed a team there and it was I think it was 150 hours a year of training. you know, three hours a week, you know, roughly. And just just a lot of training. It's like, all right, I gotta like sit here and do this on a Saturday, of course, because you're just you know back to back Monday through Friday. so how are you guys solving this problem in particular? So we got we have a really good sense of what the problem is. Yuvraj Singh (11:24.712) Wow. Wow. Yuvraj Singh (11:34.112) Yeah. Yeah. Yuvraj Singh (11:44.405) Yeah. So I think, mean, see, as you, as you said, right. Like, the training just becomes so. I would say, like such a monotonous and gargantuan task that like, enjoys building it and nobody kind of enjoys consuming it. Right. We all kind of do it like as a checkbox. and I think. That's where fundamentally it all kind of comes flat that like people, when people think about learning and development, they think about like this really boring thing that we have to kind of promote, that we have to really, you know, have it because otherwise we will get fined in tax. Whereas I think we're moving into a world where continuous learning on the job outside of the job is going to matter far more than it has in the past, right? So our view is that We don't think any organization should build and maintain trainings actually before there is a need for somebody to be trained. Right. So our future vision is that you should be able to create any kind of learning and training content just in time. Right. Now for you to be able to do that, we need to be able to run through the entire workflow completely flawlessly in a very short period of time. and making sure that the output is tailored to you. So what that means is that you can simply say, Hey, I need to be trained on XYZ. Here are two documents that I have. Build me a 30 minute video tutorial on this. Right. And Aristo automatically assesses whether that is the relevant training for this individual, depending upon the role that they are in. Do they have the right access to the right materials? If it doesn't have access to the right materials, does it need to engage its sourcing agent? It goes and fetches additional information, whether it is from company libraries, whether it is from external sources, cites it, structures the entire learning outcomes, creates the storyline, tags everything according to the internal guidelines, internal compliance requirements, makes sure that it is sent for the system, like the record of systems back to the central LND team for their approval. Yuvraj Singh (13:56.127) And ships out a full fledged, well-produced masterclass kind of video training for this individual. Right now you have this power, which existed only with learning specialists, instructional designers, and SMEs in the past distributed to every single employee. And they can create any kind of learning content, literally just in time, while still the central team, which cares about compliance, which cares about governance, which still has to report to the auditors has all that data. that who is using what information, how they're creating that content, how they're consuming that content, right? So essentially for every single action that was being taken by humans between those stages, or that were being automated by tools in each of these stages have essentially been like automated through one agentic flow. But at the right point of time, we do have like human oversight, which can basically kind of capture and steer the system into the right direction. So it's an end-to-end agent to clear. Brian Bell (14:57.464) Yeah. So effectively what it sounds like, because we're now enabled by AI and agents and video generation, you're able to get this content system, L and D CMS kind of system that's just in time, completely customized, that would would have cost you millions and milli millions of dollars before this technology existed. This is the kind of stuff Microsoft had, you know, 10 years ago, but they had a whole team of probably hundreds of people and I mean the videos were really well produced with actors and everything. I mean, and it was they're kind of fun to watch, right? It was kind of cheeky and you know, they made it kind of fun and you know, and and relevant and they made it as fun as to watch as possible, you know, like really uncomfortable situations at like after the ball game and stuff like that. But now you can do that at scale with AI. And so now it's this incredibly democratizing example, like power like. Example of the power of AI. Yuvraj Singh (15:56.883) Exactly. Exactly. And I mean, that's, that's essentially what, what we are telling that like you don't, I mean, a Microsoft could do that because they have the resources, they have those large teams, had deep pockets. Brian Bell (16:05.346) Yeah. Throw tens of millions of dollars a year, probably a hundred million dollars a year at a training at Microsoft. Hundreds of millions of dollars. Yeah. Yuvraj Singh (16:11.327) Yeah. And now you can essentially enable any small and medium sized company as well, or even a mid enterprise company, or essentially organizations who are training other companies, right? Like there are so many organizations who are in the business of content creation or in the business of like training professionals. And for them, high large volumes of content creation has typically been done by like outside consultants, right? So all of those organizations are now enabled to essentially scale up the production by 30, 40. And that would naturally happen for almost every enterprise, right? Because if you are going to give every employee the ability to create their personalized learning content, that means that it's 30, 40 X the volume that they are creating today, right? And nobody even Microsoft would not want essentially that hundred million dollar budget to like scale up to you know 300 400 or like three billion four billion dollars. So so that's the advantage you get to control the quality you get to control the governance you get to scale up the production and actually cut down the cost significantly. Brian Bell (17:18.318) Right. So what kind of questions do you usually have in that first pitch meeting? You know, who you where are you finding traction? What's your ICP right now? It sounds like it's kind of the bid market, but yeah, maybe you could walk us through some some questions you get on that first kind of demo meeting with the I don't know, who who's that typically the head of training and development or the head of HR maybe. Yuvraj Singh (17:34.773) Yeah. Yuvraj Singh (17:40.267) Yeah, think so our typical kind of like customer when we started this was essentially, I mentioned, organizations who were creating learning content for other organizations. So think of like healthcare training institutions, publishers. maybe government upscaling organizations, the large learning providers who were creating training for other enterprises because these organizations have to create like hundreds and thousands of hours every single month. So for them more high quality content, quicker to market equals to more revenue. Now we are working with lot more, like I would say mid-sized organizations and even some government organizations who are essentially, and as you mentioned, right, like these are chief HR officers, chief learning officers, the vice president of upskilling and training and development. These kind of individuals are the ones that we typically speak to. I think honestly, like in our experience, like people get completely blown by when they see the demo. We have very high conversion from demos to pilots and pilots to like rollouts, right? So between demos to pilots, we see easily about like, I would say 40, 50 % conversion and then from pilots. Brian Bell (18:55.31) That's really I. Yeah. Yuvraj Singh (18:56.596) And from pilots to like rollouts, it's about like 80, 85 % conversion, right? So our biggest challenge is we need to be in front of as many people as possible, So, exactly. Brian Bell (19:06.072) Yeah, it's top it's a top of the funnel challenge for you and and getting enough SDRs and account AEs to sort of scale and that understand how to pitch this and answer the questions. And too bad you can't have AI for that yet, but Yuvraj Singh (19:13.524) Exactly. Yuvraj Singh (19:18.124) Yeah. Yeah. Yeah. And I think, and I think honestly, like, you know, the way, the way things are like, kind of happening right now, I think just simply bombarding people with emails is not going to work. Vom connections will probably like, you know, get you much further down. And that's where like most of our customers have also kind of come through. Like we've gotten like a lot of referrals. we've gotten like some inbound leads, et cetera, from the content that we've generated. but yeah, I think coming back to, your like question around like, what is the first thing that. Yeah. Brian Bell (19:27.694) Mm. Brian Bell (19:43.982) Yeah, I wanna I wanna I wanna go I wanna go on the growth corner for a second. I need to come up with a with a funny alliteration for for growth goals or I don't know. Growth corner. growth corner. what what is working right now? I've heard that you just have to like just get in person, you know, and go to conferences and just talk to people and what what do you find is working for you as you yeah, try to grow? Yuvraj Singh (19:48.544) Sure. Yuvraj Singh (19:58.529) Growth corner it is. Yeah. Yuvraj Singh (20:08.97) Yeah. No, I think that's, I think things have changed a lot over the past couple of years. Right. And, I think conferences from being like unsexy ways or like unstructured way, think are becoming like certainly like one of the channels that you should execute properly. Like our strategy also has been like over the past, like year that let's hit like four or five main conferences. Let's. really build the qualified leads funnel from there and then let's try to like exhaust that over the next one, one and a half, two months, right? So I think that works out really well. Putting out a lot of content on LinkedIn, particularly like I think stuff which is relevant for your buyers announcing who you are working with. That also happens to be like highly useful. We've gotten some really strong inbound leads, like one large government entity, which essentially would almost never work with a startup. you know, and go through like this massive tender process, figured out a way to work with us purely on the basis of like something that we posted on LinkedIn. Right. So I think those two kind of things like have worked out the best for us. Another thing that we are really gung-ho on is that serendipitously over the past one year, we've built a very solid network of about 2,500 subject matter experts and learning specialists. And this happened completely like, you off chance in the sense that some of our clients said, Hey, we've traditionally worked with consultants. Your tool is great. We don't have enough people in the organization to work with your tool. Can you help us like bring some subject matter experts, learning specialists who can use this tool and deliver us an end to end service. Now we're a team of five. So we got, a lot of contractual instructional designers, learning specialists to come and use the platform and deliver this final work product. for our clients. That system happened to grow like quite a lot with a lot of customers. And we were able to attract more and more contractual folks. And now we have a network of about 2,500 people who are essentially users of the platform. They're earning from the platform, delivering value to our clients, and they're becoming evangelists as well. So I think one of the things that we are going to potentially really explore is how do we use this group to get more warm introductions? Because a lot of them, Yuvraj Singh (22:27.616) come from mid to senior level positions in different enterprises in different parts of the world. So using them strategically to like, you know, get our foot into the door in some of those organizations is gonna be like an interesting strategy that we deploy. Another thing which is working out well and we're in like kind of early stages of that is often your content gets outdated very quickly, particularly like how quickly things are changing today. And on top of that, If you are in a regulated industry, any compliance changes, any product changes, policy changes can make your training content like completely out of compliance. And that is a big, big risk for a lot of organizations. So we've built a system which basically can read your existing content and highlight the problems that exist. And we want to now we've tested it with a few clients and we're now going to launch it completely like for free for any organization. so that they can come and test like, you what are the issues, what are the errors in their existing materials? And then how do they fix it? How do they kind of ensure it doesn't happen in the future is through Aristo, right? So that is going to be like a PLG move. Brian Bell (23:37.868) Yeah. An analytics as the freemium model. I've I've seen that before. That's a that's a growth hack corner. really smart thing. It's like you show a little value and it's just like you have a problem, you know you have a problem. Now measure the problem. And by the way, we can solve that problem. Let me show you a demo, right? Yeah, that's really smart. Yeah. And you want it to go back to the question I had before. What was that? Yuvraj Singh (23:41.729) Yeah. Yuvraj Singh (23:47.851) Yeah. Yuvraj Singh (23:52.427) Yeah. Exactly. Yeah. Yeah. Yeah. Yeah. So that's the note. Yeah, think, see, one of the, I mean, the, lot of people that we like, you know, get to the demo stage. We thought originally that people would be super excited with the speed with which we can create. And we often said that like, you know, you can just essentially like put out a 12 hour training in like three hours or like you will save 30 eggs. Right. And I think over a period of time, what we realized is that this is not the most pressing value proposition for people. whether it is three hours or one day or two days, that doesn't make much of a shift, right? It does make a shift when you're moving away from like months to maybe like a week or a few days. But the fun and people have seen a lot of these claims with a lot of other providers. think a fundamental question that we get from people is how sure are you of the quality? Right. how accurate and compliant will your system be to our internal guidelines? And, I think that's. essentially where like we get a lot of our questions. and I think we've kind of anchored the product also to be like really, really like a dead end to organizations, qualities and specifications. Right. And, and it, is an interesting thing, right? If I tell this to like, any typical VC or like somebody who's also not in the space, they're like, yeah, but this is, you know, I mean, how hard could it be? How difficult could this be? Like imagine every single organization and they are extremely particular about the degree with which the corners should be circular, right? The kind of icon to use, the kind of transition not to use, the kind of like super small, tiny regulation and code that has to be like compliant. Everybody has their own kind of set of like specific regulations and nobody wants to budge from it. And if you're talking about content that is going to be reflective of your organization's core values, Yuvraj Singh (25:52.895) You don't want general genetic average summaries being produced. You want your unique fingerprint on that, right? So a lot of the questions is that how much control do I have on what I produce versus me just trusting your system and you producing something, you know, which will sound same for like organization A versus organization B. Brian Bell (26:18.382) Sounds you got something here. what are you excited about for the future? Yuvraj Singh (26:23.98) I think what I'm excited about the most in the future is essentially that I think for the longest time there we tend to think that like you know only few people have the right kind of knowledge and only few people and rest of the world is essentially just kind of like trudging along. I think when you look into the enterprises knowledge kind of exists with a lot of people. Either there is no mechanism to extract the knowledge out from every single organization's exit interviews don't really work out to be well. SOPs don't really like, you know, are properly utilized or they're not like kind of properly capturing the essence of the knowledge, the context of that situation, et cetera. And nobody really frankly has the time, right? What I'm excited about is that there is insane amount of knowledge that is trapped in every single employee's mind. so one is that, yes, they will have the ability to build something just in time. But the second thing is they will also be able to literally give us a raw dump of what they know, what they don't Combine that with some existing SOPs or institutional knowledge and the crazy abundance of interesting, insightful pieces of, you know, new insights can kind of come new knowledge can come out of that. Right. So. We don't just think that like you will be able to create just in time knowledge. will actually be sourcing just in time knowledge from literally any and every employee across organizations. Right. So I think that's why we think that like we have the potential of being like the operating software, OS layer for every enterprise and it's knowledge base. Brian Bell (28:15.864) Yeah, so it's almost like you get the you know, a lot of people learn best by apprenticing, right? You kind of sit with somebody who's more senior, right? They've been there for years, they've been doing the role, and now you just go shadow them for a while and kind of learn the ropes. And that's how a lot of people learn. But it sounds like the vision in the future is not only can the L and D department, or whoever's in charge of training, you know, source and create all this material that's compliant and deliver just in time and all that. Yuvraj Singh (28:24.491) Yeah. Yuvraj Singh (28:31.83) Yeah. Brian Bell (28:45.24) But you can also source that actual material from all the digital footprint and and and literally the heads of of the people that know the the information. I I recall this actually as a as a as a product manager, many, many times, you know, the the the training guys would come in and talk to me. They just want to know like, okay, what's going on with this product and like who's it for and what's it do? And and give me a demo and You know, so I'm training the trainers a lot, right? So it's just it they're trying to get it out of my head into a structured program. and that's part of what you guys do, is you're kind of collecting all that knowledge, synthesizing it and creating programs. Yuvraj Singh (29:21.205) Exactly. Yuvraj Singh (29:28.426) Yeah. And I think it's very interesting that like what we're seeing with a lot of our customers today, like we get like these raw video files, right? Of like zoom recordings or like people literally just talking to the camera. And oftentimes what happens is like, I mean, we're not like great public speakers in general. Like, and particularly when we're trying to like teach someone something which could be boring or monotonous, we're not the best storytellers. So we will get like these Brian Bell (29:38.35) Mm. Yuvraj Singh (29:54.551) two hour long video recordings of where the person is mumbling or they are going over the stuff again and again, or they go into like a random corner, which might not be the most relevant or then, then they give you like something super insightful, which might be relevant for a different topic. Right. And literally just being able to take that raw dump, right. Process it through a Rostos pipeline and give you like five or six, like amazing five minute, 10 minute, 20 minute, like solid instructional videos. with proper production, proper kind of visualization of the systems, digital walkthroughs, animations, et cetera, so that you're not just listening to a person when you're consuming that, but you're actually kind of being able to like visualize this and like solidify your understanding as well, right? And that being logged into your knowledge system. So tomorrow when you're accessing it, you're not just kind of going over like, you know, random video files or random transcripts. have full-fledged polished like outputs which can be utilized in multiple different manners. Brian Bell (30:56.75) Incredible. I actually I was just thinking of a friend that used to sell content management into the enterprise. This is probably like ten or fifteen years ago. really interesting. Brian Bell (31:12.792) How do you how do you how do you know people are retaining the information, right? So you're trying to train these employees. Are you like doing like little pop quizzes at the end of the videos, checking in a week later? Like what's the best pedagogical practice there? Yuvraj Singh (31:26.944) Yeah. Yeah. So I think typically like what, what has generally always happened is that like LMS is track a lot of information, right? and the problem is that like the track the information and they represent that in like nice crafts. Right. Yes. But if you can't do anything with that information, it's only good for like reporting, right? So we, we, any asset that you've built with us though. Brian Bell (31:41.654) Learning management system, just in case anybody's listening at this point doesn't know what that is. Yeah. Yuvraj Singh (31:55.583) essentially tracks all that information, no matter the elements that it is being consumed at. Right. So it's not just that like who completed it and who scored what in a specific quiz. It even captures, what was your engagement level? Right. At what point of time in a video or in a reading material or in a sandbox environment, did you lose interest? Did you stop your activity? Right. There is a chat agent. to whom you can ask all the questions and it has essentially all the information through which the entire asset was built. So what kind of questions are you asking? Are they first degree? Are they second degree? What's the depth of the information that you're kind of getting at, right? So we track a lot of this data and not only do we report that, we actually like kind of use that again to improve the creation process in the future. One of the things that we are seeing with a lot of organizations today is that... passive assessments of like a quiz at the end of a video or like in the middle of a video are becoming not very exciting for people and people are moving away from that. Right. So one of the things that we asked, we get often asked is that, okay, I'm teaching someone, how do you build like a budget for a specific project? Right. Build me a sand project, build me like a sandbox project where they can actually like, you know, create the budget and kind of calculate the ROI and calculate like, you know, what the optics would be and how they would allocate that across various initiatives and resources. Now, nobody is telling a Rusta that I need to build this. All we have is the learning objectives and the raw content. It automatically decides that to achieve these learning outcomes, I have to first establish the foundations using this video and some reading material and case studies. Then I'm going to give the person a real hands-on activity and experience in a sandbox environment with a proper UI. And how they engage with that is where I will actually know that whether they have understood that concept or not. Right. So that's, that's how we are. We're kind of like tracking how people are learning this content. Brian Bell (33:56.35) So put yourself out five years from now, what disappears from corporate learning environments? Yuvraj Singh (34:02.312) I don't, I think my, my, my kind of point of view is LMS is as we know them today, will completely shape, shape, shape shift. I don't think you will need to log in into any environment, and click on a course to consume that. It's going to be much more conversational. It might even be like a layer with your large LLMs. where you're simply just saying, hey, I want to learn XYZ stuff and stuff gets built on the spot, gets sourced to you, your activity is being tracked. And that can potentially be recorded in some mutated form of an LMS or it can go back to the creation system like ours. But I don't think that the entire user interface or the entire environment, how learning is consumed today in an LMS. will exist in the future. Brian Bell (35:03.574) You said that knowledge expires quickly, especially you you know, regulations changing, AI is obviously changing really rapidly right now. What types of organizational knowledge decay the fastest? Yuvraj Singh (35:16.63) Hmm. I think that's quite interesting. would say, I think nowadays, any kind of knowledge-based, procedural, skill-based knowledge is kind of decaying the quickest, right? If you thought that the best way of doing an analysis was you have to assess it over these four frameworks. I don't know, like, or you would use like a specific tool. So actually, yeah, let me, let me reframe that. Right. So I think skill-based knowledge expires quickly. And that skill-based knowledge, which is specially reliant on not just like theoretical frameworks, but like usage of tools. Right. So if I'm really good at using XYZ tool, I think six months down the line, there is going to be like a better version of that tool. Right. and it might allow you to do your job much more quickly. So I think that is something where you can't have trainings built on old tools with old features, old versions, be something that you can really rely on. I think the other thing is market context. That is something which typically is really undervalued, and in most organizations, people are actually not taught. on how the market that they are operating in is shaping up to be. Whether you think about sales roles, whether you think about consultant or customer success roles. And given how fast innovation is happening today, I think even that is essentially like expiring like almost on an every month basis. Brian Bell (36:56.748) If learning becomes just in time and then just in this automate automated process, does the word training become obsolete and need what do we and what do we call it instead? Yuvraj Singh (37:10.1) That's that's that's a really, I think it's preparation. it's a, it's like you, you prepare for an interview, you prepare for your next project, right? it's kind of like transforms into preparation. Yeah. Yeah. Something like that. Brian Bell (37:20.462) Hmm. Yeah. Like task and skill prep prep or something. Yeah. Like I need to do this thing I've never done before. I'll just grab this like just in time system that like synthesizes and distills all the all the knowledge that we have internally and externally from approved sources. It's compliant. It's like, how do I do this thing? You know, you know, five minute video. Yeah. Just preps me right away. Or maybe it even just coaches me, right? Yuvraj Singh (37:38.314) Yeah. Yeah. Yeah. Get me ready for that. Exactly. Yeah. Yeah. Exactly. Yeah. Yeah, yeah, yeah, I think that would probably. Brian Bell (37:49.166) That'd be another kind of extension of the product eventually. Yuvraj Singh (37:52.333) Yeah, you probably will take preparation as like kind of laying the foundation. Of course, you're not going to grasp everything just in preparation. So when you move into actually doing the task, you'll probably have like some sort of agent which is watching over your shoulder and kind of assessing that whether you're actually applying what you've learned or like where are you making misses. And it can give you like real time feedback, right? And particularly if that is happening in like some sort of a digital environment, which is constantly monitoring you. Yeah. Brian Bell (38:13.976) Yeah. Brian Bell (38:20.802) What are you seeing in terms of human expertise become more valuable in an AI driven world? Yuvraj Singh (38:28.94) Yeah, I think I, so I heard this phrase from someone recently, which I think they put it really well, that what you're most frustrated with when AI does that task, that is what you are really good at. And I think it's very difficult for people to really kind of like extract that specific task. or that specific action out of their entire job description. Right. So typically like when you think of like AI solutions and you try to go and like, you know, sell that to people, they think you're just automating their entire job and they're to be out of that job. Right. But actually you're trying to automate a lot of the mundane repetitive tasks, which can be predictably automated for them. But a lot of the thing where more like I would say more like discernible judgment is required. That is where your expertise really kind of shine. So I would say like from a procedural process-oriented tasks, that's where the human expertise really comes into play as to why certain decisions when they're not factually wrong, but might be wrong for your context or your organization have to be taken into account. I think that's where an expert's opinion matters a lot. The second aspect is I think right now there is still some level of like fundamental mistrust with AI systems. And that's because they do make mistakes sometimes, right? So when it comes to like quality assurance, when it comes to like kind of proofing that like this is ready to go in front of people, that's where again, like we see like a lot of reliance or like people feeling comfortable that we would want a subject matter expert to review this, right? Whether it is internal or whether it is external, we would want them to kind of sign off, right? Which has been basically one of the reasons that this entire SME network was developed around our product as well, where people love that a third person has actually reviewed what was produced or was monitoring the entire process of creation. Yuvraj Singh (40:46.908) Now, one of the good things about how we built the system is that we're almost capturing every single action, decision, review, comment that takes place through the different stages. So anything that we mess up on, the subject matter expert is essentially telling Arusto that, hey, this is what you should have done for this specific context. That becomes the data. that we then train our models back on so that next time around we don't make that mistake, right? Whether it's a formatting mistake, whether it is a sourcing mistake, a content mistake, whatever it might be. But yeah, I think these two spaces are where expertise still matters the most. Brian Bell (41:30.712) Let's do some wrap up questions as we kind of wind down the podcast here. what is a private rule you live by that has served you well? Yuvraj Singh (41:33.558) sure. Yuvraj Singh (41:43.038) A private truth. Yuvraj Singh (41:47.894) Hmm. Yuvraj Singh (41:52.095) I think I kind of do believe in Murphy's law a little bit that like if I think something has a probability of happening and if I particularly think that it's going to be negative, it's likely going to happen. So I try to like prepare for that worst case scenario for sure. I don't apply the same rule for like something positive. if I think that like something shit is gonna hit the fan, it likely will hit the fan. So let's let's like kind of resolve that, right? Yeah, Okay, that's interesting. Brian Bell (42:21.986) Yeah. Yeah. I t I tend to think almost like the opposite of that, which is like if I focus on the rock on the road, I th I'll probably hit it. And so best not to like focus on the rock at all and I'll solve it if I need to solve it and not worry about it. Cause if I worry about it, I'm almost like attracting like the universal energy towards me of the negative, right? It's a little manifestation energy thing, but I tend to kind of trust try to focus on the positive. Yeah. Yuvraj Singh (42:35.659) Yeah. Yuvraj Singh (42:40.074) Yeah. Yeah. That's, I think that's probably a, that's probably a healthy, that's probably a healthier mindset. I think because yeah, this, does lead to like me hitting the panic button a few times when it was not necessary. Right. So yeah. Brian Bell (42:53.186) Yeah. Brian Bell (42:57.41) Yeah. Yeah. And then you usually things have a way of like working themselves out, like too, right? Like it like sometimes you can just ignore a problem and it'll go away. That's why people don't go to the doctor, I guess, when they're sick. They're like, I'll just like go w work itself out. And like sometimes you like really need to go to the doctor and like check it out. anyway. Yuvraj Singh (43:04.225) Yeah. Yuvraj Singh (43:09.236) Yeah. Yuvraj Singh (43:15.808) That's very interesting. My co-founder has that mindset, right? So I think we kind of offset each other pretty well, right? Like he's very calm, level-headed most of the times, doesn't stress out, which sometimes stresses me out, that why are you not stressing out? But I think, it kind of works out. Brian Bell (43:20.915) interesting. Okay. Yeah. Brian Bell (43:29.708) Yeah. That's funny. I I've heard I've heard that co founder dynamic many, many times actually. And that that's a that's a healthy you gotta have somebody that who's like, paranoid a little bit. and the other person's like, no, it's gonna be okay. You know, and then there's some sort of like ca compromise there between everything's like the sky's falling and no, it's all gonna be serene and fine, you know. Yuvraj Singh (43:35.628) Yeah. Yuvraj Singh (43:40.042) Yeah. Yeah. Yuvraj Singh (43:48.149) Yeah, between the both. Yuvraj Singh (43:52.072) Right, right. Yeah, yeah, makes sense. Yeah, that's interesting. Is it? Brian Bell (43:55.224) What is the most expensive lesson you've learned? Yuvraj Singh (43:58.285) I think I would say the most expensive lesson that I've learned is last year working with in our early phases, like working with certain organizations and clients in a specific geography where there were large organizations, large clients, global footprint. And they were pushing us into a direction where we probably shouldn't have as a product. And I think just because when you're in early stages, you're hungry for any and every kind of revenue, building certain feature sets for them, going too deep into building certain things for them and then not really seeing that provide any return in any manner and like, you know, some bad revenue as well. I think that was essentially the most expensive mistake because we lost like very important time. We, of course, like, you know, push the product in direction that we shouldn't have. So I think that is now a lesson for me that, you know, one, we only build something when somebody is actually paying for it. And second, we have our opinion. as to how the product needs to shape up as well, because not every organization that we work with is ready to actually adopt a product. And some might be more forward-looking and others might actually push us back into a direction where we know the world is not actually going to end up at. Brian Bell (45:20.728) Something you changed your mind on, maybe something you believed or didn't believe five or ten years ago that you'd now argue the opposite of. Yuvraj Singh (45:33.869) I would say that traditionally I would have always believed that there are certain qualifiers which really allow you to say that a specific kind of talent and people come if they have those qualifiers, if they go to certain universities, if they've worked at certain companies. But I think over the past few years, having worked with a lot of diverse set of people, I don't think that is something sure shot. Like you can, like, I think there are some folks in our team as well who are by far the most phenomenal people that I've ever worked with. And they don't come from the top schools. They don't come from the top 10, 20 schools. They're not worked with the top firms, but they are, they have great ownership. have insane amount of insights into what they're doing. They're extremely passionate and hardworking. And I think five years, 10 years ago, I would have probably said that, think a qualifier is that like if somebody has gone through some of these top schools and institutions, but I don't think that's true at all. Like, you know, we all kind of dealt certain kind of cuts and we got to make the most out of that. Brian Bell (46:41.87) Absolutely. Yeah. Yeah. I I mean I didn't get into I I went and got an MBA at Berkeley in my mid thirties, you know, and I went to like, you know, just UW, which is a good state school. and but if you would have looked at my resume at thirty, you know, thirty two, thirty three, you'd be like, you guys are loser. Like literally, like I hadn't done anything. Like really. Yeah, I worked here, worked there and traveled around, but I think everybody has a their own journey and path to what they're excellent at. And Yuvraj Singh (46:51.916) Yeah. Yeah. Yuvraj Singh (47:01.772) you Yeah. Yuvraj Singh (47:11.38) Mm-hmm. Yeah. Brian Bell (47:14.086) It takes a while to find that thing. And and I think also a lot of people who went to top schools were really, really like pushed there by their parents. and some people are just naturally gifted and they just they score high and they have like a four point whatever GPA and they, you know, they just go for it. And but there's also some people it just takes a while for them. Might take them their entire twenties, maybe most of their thirties to like really find out what they're really good at, what the world needs, what they really like to do. Yuvraj Singh (47:22.826) Yeah, for sure. Yuvraj Singh (47:29.694) Yeah. Yeah. Brian Bell (47:44.428) I mean, I I tell the story all the time. Like I didn't, you know, set out to be a VC. Like I just kind of iterated my way towards it over like a 20 year career. You know, I'm in my mid forties now. but you know, it was a lot of trial and error. A lot of trial and error. and maybe I'm not the best VC, but I really like it, right? and I think, yeah, I and I didn't come from the most spectacular background or had the best resume, right? Yuvraj Singh (47:52.544) Mm-hmm. Mm-hmm. Yuvraj Singh (48:01.035) Yeah. Yeah. Yuvraj Singh (48:10.826) Yeah, yeah. Yeah, I mean, at some point in time you realize, right, like all these labels, like all these kind of qualifiers are fictitious. I mean, sure, there is a selection bias as well, right? And the systems are not perfect through which the selection happens. And I think, sure, they still might work on averages, but at the same point of time, I think broadly you can't really like discount like people outside of it. And like, I think a lot of like traditional enterprises do that. And it's very interesting that like, actually, I think feel that people who have not been in these top schools have more of a grit sometimes. And, you know, they will go after the problem far more passionately than somebody who was like 300 other options or like, you know, think that like there is a chip on their shoulder just because they went to a certain school. Brian Bell (49:02.196) No. Well, I've really enjoyed the conversation. it's always fun to have a founder come on and I'm like, Yeah, I I back that founder, you know. it's it's always fun. Where can people find you guys online? Yuvraj Singh (49:14.124) It's on www.arusto.ai. That's about it. Reach out to me and I'll be happy to. A R U S, sorry, A R U S T O dot A I, A R U S T O dot A I. Yes. Great. Thank you so much, Brian. Brian Bell (49:21.216) A R U S T O dot com Brian Bell (49:28.28) Dot AI. Got it. Well, thanks for coming on, really appreciate it.
