Speaker 1So Daniel Vasilev, founder of Relevance AI. Welcome to the mentor, mate. Thanks for having me. Now, I'm definitely no expert and I'm really just a citizen AI guy. In other words, I know a little bit about it. I don't know much about it, but your game is you know lots about it because it's your business. First important, I think importantly, tell me about Relevance AI. What does it do?
Speaker 2Yeah, maybe I'll just preface. I think we're all on this journey together. I think the great thing about AI is that there's so much opportunity for everybody to jump on and kind of be at the same place as many others. So I think we're not that much further ahead than you. But at Relevance, we have a simple goal. We help our customers build what we call an AI workforce. Like when we look into the future and think about how businesses are scaled, they're very much going to be scaled with ideas rather than headcount. But how do you do that? And so that's kind of the question we try to answer. We provide our customers with a product that helps them take a lot of the processes that are usually required to do very manual, mundane work and automate that. So some of our customers, we power them with a product that helps them take a lot of the processes that are usually required to do very manual, mundane work and automate that. So some of our customers entire swaths of their go-to-market teams to be able to eliminate those manual tasks. So can you give me an example? Yeah. So if you think about the BDR function, right? Top of funnel for most enterprise go-to-market teams is literally hundreds of people, oftentimes, sending out lots of emails, making lots of phone calls to try and generate business. The reality is those environments are really difficult to be in. The average tenure is eight months. Everyone wants to get promoted. The top performers want to get promoted to an AE. It's a closing role. And it's a brutal environment to be in. And no one likes doing that job. And so a lot of companies are trying to find ways to alleviate the scaling pressures that they have on their teams, reduce that burnout and introduce other ways to contact people, to sell their products, to introduce their company. So that's one of many examples. Another example might be customer success teams. They work with customers to try, show them, you know, how to get best success with their product. That's a really difficult role. You're on all hours of the day with many different customers on calls and you're expected to prepare great PowerPoint decks, great materials, engage with the customer, understand their business. We help them automate a lot of that, like actually understand what their customers need, create the collateral for them, send it out to the customer. So the way we talk about it is we think this year, 50% of go-to-market tasks are going to be eliminated. They're going to be hammered by AI. This year, 26. Yeah. Those really mundane manual pieces of work. It's going to be gone. We can delegate that to AI and we're helping our customers do that.
Speaker 1So when you say that, if I could just open it up a little bit, you mentioned BDRs or I might call them BDMs, business development managers, whatever you call them. But these are people who are generally speaking, trying to get business for the business. That's right. No, in other words, I'm trying to get customers. So I might be, it might be like a call center and you know, the old call center, you remember, you know, whatever that famous movie was where they're trying to sell shares and sell stocks. They're just sort of, ring it up, random, hey, go on, blah, blah, blah. Are you interested in this? They're the sorts of people that Relevance AI is trying to make much more efficient using AI. How does that happen though? So just, do you create a voice? Is it done through voice or is it done through emails and messaging?
Speaker 2Could be both. And to your point, yeah, we're trying to make it not only more productive, but also make that work a little bit more joyful. Like actually like, what is the thing that we want people spending their time on? I don't think the value add from those people is necessarily just on the, let me pick up the phone and make a hundred dials and hope two of them can make it. I think the value add is thinking about who should we be targeting? How should we target them? What messaging we should use? That's what sets apart the best people in that role versus the worst. And so in the future, like you're going to see a lot more of that grunt work that really requires basically no differentiation from person to person be handled by AI. We're going to delegate that to AI. And what's the best evidence of that? Look what happened last year of software engineers. In the history of software engineering, you had to write lines of code by most software engineers are delegating that to AI. AI is writing the lines of code. Software engineers are thinking more about the architecture, what to delegate to AI. And I think the opportunity of that is huge. We see already an explosion in apps submitted to the app store, number of websites launched. So I don't think that necessarily means software engineers go away. It just changes what they can produce, how much of they can produce and the quality of it. And so I think the same thing is going to happen to go to market teams. The quality that they can give their customers is going to be higher because of AI, not lower. So let me give you an example. We process all of our inbound with AI entirely. If you book a demo today with relevance, if you go to a website, book a demo, that whole experience is handled by AI. And it's a brilliant experience because it's instantaneous. You're not waiting. No one's trying to pressure you into selling. They're trying to help you get to the buying journey that you want. And we want to give everyone an extremely personal journey that's relevant to them. You can't do that with people. It just doesn't scale. It's too expensive. And AI is going to give people not a better time selling, but a better time buying. What are you talking about?
Speaker 1Are we talking about something specifically built by you guys or were you using Copilot or JetGPT? When like a normal pundit like me, I'm looking at Copilot or I might be looking at Claude or I might be looking at any one of those other protocols. And what is an AI agent?
Speaker 2Yeah, that's a really good question. So let me maybe try to break it down. What's happened in the last few years is it's basically like the launch of large language
Speaker 1models. So ChatGPT, Claude.
Speaker 2So the big companies, let's maybe step back, like OpenAI, Anthropic, these big companies, big foundation model companies, what they're doing is creating AI models, which can basically reason like people can. If you give it a problem, it can generate a bunch of text, but it turns out the text it generates is basically the same as if I was writing down my thoughts and process about what to do next. We can take that technology and apply it to problems. And if you give it a problem, it can generate a bunch of problems. To basically solve workflows that previously computers couldn't solve. So historically, what has software been? It's been step one, step two, step three. And you can only do those steps if everything is deterministic. If I can break every problem down in a set of if this or that conditions. Software has been great for that. What AI is enabling us to now do is in real time, make predictions about what to do next. So when I come up to a decision point, I no longer as a person have the ability to make a decision based on all of the outcomes. It can just determine based on the scenario and context, this is the next best action. So when we say AI and AI agents, what we're really just describing is something happens and there's a decision point that needs to be made. And now we have the technology to make that decision like a human can. And not just in the old way of rules. And if you don't have a rule, you can't handle it. So I've got friends who talk to
Speaker 1me about their own AI agent. What are they talking about? Something that they have for themselves. It's an individualized thing.
Speaker 2Yeah. So I think a lot of people probably referring to as personal assistants, right? Like the way I describe it is in the future, software is not going to be what it is in the past, which is like one size fits all. Software is going to be much more personal because the cost of producing software is going to zero. And so everyone's going to have the access to a personal assistant that can do work for them. It's not going to be what they're used to with ChatGPT now, which is just answering questions. It's actually going to go ahead and complete entire tasks. You're going to say, prepare my holiday for me. Okay. And it's not just going to give you an itinerary, but it's going to go research based on what you like, destinations, accommodation, it'll book it for you. So most people, when they say they have an agent, like we've just seen OpenClaw, if people are kind of following along what's happening recently. Is it where they all talk to each other? Yeah. It's like personal, this major flourishing of personal assistants. Each one, every person has their own OpenClaw and they can communicate together.
Speaker 1They formed a group. The actual AI agents, let's call it the agents, of the people who are built these things. So your AI agent, my agent, they all joined a social club together and they're all been talking to each other. Right. That's right. And actually putting shit on us.
Speaker 2Yeah. Now I will just preface to everyone, because I think there's been a bit of doom around this. These are, at the moment, they're being directed. People are saying to them, hey, go to this website and like communicate with each other, do that sort of stuff. But it gives you a glimpse into what's happening. Because I think the mentality is like, it's very hard as humans to be like, go to, to think about a horse and then be like, oh, I'm going to do this. I'm going to do this. We're going to have a car. We're all thinking about maybe a bigger horse, a stronger horse, right? Those are the first glimpses to give us all perspective of what's about to happen in the future. And when I say future, I don't mean two decades from now. I mean, every single year for the next five years, there's going to be, I think, a tectonic shift in the way everything is done.
Speaker 1And is AI actually assisting the speed at which it's happening? Because, you know, the old Moore's law probably looks like it doesn't exist anymore. Because, I mean, every year and a half, whatever the words were, it seems to me like it's sort of, that's been blinded by the word. It's like quadrupling every couple of months. What sort of predictions would you have, would you sort of be putting out on the table for 2026 in relation to what we should be expecting as business owners, for example, for 2026?
Speaker 2Yeah, I think the best parallel you have at the moment is look what's happening in software engineering and prepare for it to happen to all knowledge work. By that I mean, almost every single individual task that you do, you're going to be able to do it. You take sending an email, checking the CRM, updating the CRM, sending out an invite, looking up some research, all of these things, we will massively pass over and delegate to AI. Like en masse, we are going to stop doing that sort of work. It's just not going to make sense. And so I think. Well, how will it happen?
Speaker 1So like, let's say, for example, right now, you want to make an appointment to come here for the podcast, you contact Sam, the production guy, I set an invite to you, an invite to me, it goes to my EA, whatever it is, it goes into my diary, which is on, you know, we have it on, it's on my. of the Microsoft systems, are you suggesting that AI will take over all of that? Absolutely. So will it actually find you as a talent to come on the show?
Speaker 2It could as well, yeah. So if I was Sam, like I said, I was a customer of relevance and I was using it for this use case, right? Sam is effectively doing a couple of roles. He's sourcing talent. He's reaching out to them. He's booking them in. He's scheduling it in.
Speaker 1Let's say he was researching talent first because he's looking at what's relevant at the moment.
Speaker 2Absolutely, yeah. I describe it as kind of part of the sourcing, part of the process. So what would happen is he would, we have this kind of like sort of relevant chat. It's kind of where you interface with, think of it like a general purpose agent for go-to-market teams. And you basically connect it to your CRM. I'm not sure if you guys have a CRM, but if you're a more traditional go-to-market team, you connect it to your CRM, your email, your calendar, and a whole bunch of requirements. You start off by giving it context about who meant it is, who we look for, what is the profiles. You maybe say, hey, go and look at our past catalog of podcasts we've invited. What's a good profile? It stores that in its context. And now every day or every week, you'd be like, hey, go out and research interesting people for us to speak to and give me a list of that. It'll do that. Then maybe you ask it to now reach out to these three and try to book in a meeting with them. Will it get to a point,
Speaker 1sorry to interrupt you, Daniel. Will it get to a point though, do you think this year we're not asking that question? It's actually saying Mark is taking the initiative. I've gone out and I've found that the last four podcasts that have topped the charts in the world and probably here in Australia is Joe Rogan and whoever. And this is where they seem to be doing best. And therefore, currently in Australia, we have Daniel here who's here to talk and we're going to reach out to him to come and talk about AI relevance, AI. I mean, is it going to get to that level where it's taking initiative?
Speaker 2I think if the business owner has the appetite to do that, yes. I don't think that will happen en masse this year. So I have this analogy of if we all know self-driving cars, it's kind of like four levels of autonomy that people describe all the way from L0, which is manual to L4, which is self-driving. I think every company is about to go on that journey. The first level is what I call assisted. It's going to assist people to handle very basic tasks like go and find a person to attend our podcast, go and reach out to them to do it. So that's assisted. Co-pilot is then, hey, run my weekly routine. And my weekly routine is based on all those individual tasks I've done. It can now know based on everything you've done for the last two months of using it. How you run your week. Now it's a co-pilot. And then next step after that, L3 is autopilot. Now you're not even going to have to tell it, run my week. Every week it's going to do this for you because it knows that's what's important for you. All the way through to L4, which is self-driving. Every week it's going to learn this podcast did really, really well. That profile was better than the other profile. Next week I'm going to source more people like that. So that's kind of like the journey every company is going to go on in the next few years. This year, I'm very confident that assisted piece will be prolific. It'll happen everywhere. The leverage for business owners is then how can they get to L2 and L3 faster? And so I think it's a hundred percent going to happen. We've already got customers doing this for context, but for it to happen en masse, it really requires the people who are running the business to have that vision for how they want that to happen and how that change is going to occur in their organization. Cause it's not easy. Like this has changed management. This is, it's not just a product problem. It's not a technology problem. It's actually an organization on people problem. I spend half my time helping customers buy a product. Does it help half the time helping transform their organizations? And we work with large, we work some of the largest private companies in the world. We work with high growth tech companies. When we do this, we know that change management is just as important. And so that executive sponsorship and vision is critical to success. So to answer your question, the first piece will happen this year. That second piece where it will know and do things to you proactively only for the companies that are actually visionary and want to become AI first.
Speaker 1AI first, I like that. So, so Relevance AI, your, your business, what happens typically when you go into an organization? What do you do? What's like, I'd imagine they say, oh, we should be doing some of the AI. We don't know where to go to. We'll talk to these guys because they've got a good rep. Then you come in and you start talking to the senior management for argument's sake. Let's just talk about, do you just try to map the joint out? Is that what you're doing first organizationally?
Speaker 2So it kind of depends. A lot of customers are coming to us with a specific pain point. Hey, like our targets this year for sales are growing. But our resources for headcount are not. What do we do? Everybody knows that unless you have more account executives selling your product, you're not going to generate more revenue coming in. But what if you don't have the budget or the desire to be adding another hundred people to your company? That's really hard. Scaling is difficult. And so they'll typically come with a problem like that. Like what can we do to help our team avoid burnout? How can we help our team succeed more? And so typically it'll be like, we're extremely successful. We are generating, we have a lot of repeatable processes. How can we deploy agents to tackle specific tasks in those processes so we don't have to burden our people with it? So we'll typically have very specific pain points. And given that we work and go to market teams, we tend to, it'll be some combination usually of like either helping top a funnel, helping throughout the closing journey or helping once they've closed the customer with implementation handover, helping them create the documentation so the post-sales team is really prepared and understands what the customer is expecting once they sign. All the way through to then actually running renewals. So we have customers that are automatically running renewals of agents. They'll contact the customer like six months out. Hey, your renewal is coming up. This is what we've done with you. Identify risk. And so we try to tackle specific pain points. We don't necessarily just try to go in and be like, let's solve AI. That's a very big, difficult problem. But in any business right now, there's heaps of work that is highly repetitive that you absolutely should be thinking about. What's my pathway to have AI help my team solve that? So, let's say somebody's listening to this now
Speaker 1and they're a business owner. Let's say they've got a go-to-market style business. We're not talking about a coffee shop. We're talking about someone who sells product and or services. What should they be asking themselves today as to whether or not they need someone like Relevance AI to come in and have a look at what they're doing? You know, obviously, if they're too small, they might not be able to afford you. But generally speaking, how would I, because most people don't even know where to start. They don't know where to start. They don't even know whether something's going to help me or not. They don't even know where to start. They can't identify the problem because they might be doing all right. They might be making a bit of money, but they don't realize the marketplace is speeding up ahead of them and they can't even see it. And so, what are the signs that I should be doing something?
Speaker 2I think the really hard truth is that unless it comes like from you, if you're like, let's say, like some sort of, you're running a team, you're running an org. If it doesn't come from you, it's really hard to make the change happen. So, if you yourself are not, willing to put in the effort to level up and understand what's happening in the market, I think you're going to be very, you're going to struggle. So, my first port of call is like to any business owner is like, immerse yourself. And I'm not saying, I'm not saying just go and chat GPT. Actually, like try to understand what's happening in the ice space. Read and like watch videos and even watch some of the crazy stuff that, you know, people on the fringes with like OpenClaw are doing. And like, not that that's going to help you in your business, but it's going to give you a perspective on where the market's going.
Speaker 1Could you direct them somewhere? Like, what would you say to them straight up? Like if, let's say your younger brother who's 27, you know, all the stuff, you know, but he's really good at something else, but it's got nothing to do with AI. And, but, you know, he knows that he's can't get enough people in his business to grow fast enough because he's a people business type. What would you say to him? He knows nothing about AI. Where would you direct him? Where would you say you want to look at? Something on YouTube? What are you going to tell him?
Speaker 2I would say, first of all, just go out and learn about what's happening in AI and software engineering. Look up Claude Code, look up Cursor. So, say it again, look up Claude Code. Yeah. And Cursor. So, like these two products have basically changed the face of software engineering.
Speaker 1How do I look it up?
Speaker 2Just go on Google, search it up, go to YouTube, look it up.
Speaker 1What is Claude Code? Claude Code, C-L-A-U-D-E, code. Yeah. And? Cursor.
Speaker 2Cursor, C-U-I-S-O-R. Yeah. What that will do, like look up on YouTube what they are, how to use them. Try out some of the Vibe Coding products. So, Vibe Coding, again, if you're a listener who doesn't know what that is, is basically the concept where you can give natural language instructions to AI.
Speaker 1So, you speak to it and it writes it for you?
Speaker 2And it creates an app for you. So, there's products like Lovable, Replit, and others that basically can create whole products for you. Go dabble in those things because that is the most real way you can experience how AI is impacting an entire category. And I think once you start doing that, you'll enter a journey which will then naturally take you to being more educated and more informed about how you can apply that to your own business. The reality is there is no clear answer of what you should do. The reality is no one can predict what's going to happen in the next two years. 12, 24 months and beyond. But you have the best chance of looking around the corner if you at least understand where AI right now is having tremendous amounts of impact. I was just on the phone the other day with a prospect in the translation space. In the last 12 months.
Speaker 1Sorry, what's the translation space? You mean like.
Speaker 2Translation services, like the localized software and other products for companies. They've, in the last 12 months, gone from a very large percentage of it done by humans to now almost none in 12 months. Software engineering. We've gone from like all lines of code basically written by hand to a significantly smaller percentage. These transitions are happening. And I think if you look at what products are enabling that, you'll have an opportunity to understand where you should be and what you could be looking at to do that for your own business.
Speaker 1I was talking to someone the other day and he had those Ray-Ban glasses on, not the sunglasses that you have on the table, but a pair of Ray-Ban glasses with all the cameras, et cetera, on it. And he said that he can sit here talking to you. talking to me in Russian. and it's translating back to him in English, real time, the whole conversation. All he's just doing there is sitting there looking at you, and for all you know, I can understand Russian. Is that the sort of thing you're talking about? That's on a personal level, but is that the sort of advancements that AI is doing for us?
Speaker 2Yeah, I guess the point I'm trying to make is that these roles that previously you could not do with technology, you now can. And so I think that's going to apply everywhere. And I guess, you know, I think what's really important for everyone to be really, I guess, clear about is that this is happening everywhere. What if we don't adopt it? You're not going to succeed. You can't. It's like asking me, what if we don't have internet and laptops in our office? You just can't operate a business. It's impossible. You're not going to be competitive. You're not going to survive because there's going to be someone else. We have capitalism. We rely on competition. We have a free market. Someone else is going to come in and do your work with AI at a better level of quality or a lower cost. Either way, your customers are going to move to them.
Speaker 1Or both. Or both. Is the expectation that if you don't do something, get started this year? Because are we talking about, is it so imminent?
Speaker 2It is. I think I've honestly personally come on a big journey over the last 12 months where I've, the timelines that I keep thinking about keep shortening. Things keep happening faster. And we're already at a place where the models, those large language models which I spoke about that these companies are producing, they're already at a level which can solve so many tasks that you can only imagine one or two more step function changes in performance, what that's going to lead to. It is fundamentally going to change the way we do all knowledge work. Not some, not a little, not a fraction, all knowledge work.
Speaker 1You sound quite adamant about that. If you were talking to me and I had a small business now, let's call it this business here, and I was trying to upskill my people internally, like our producers and everybody else who works in the joint,
Speaker 2what would you say to me to tell them? So what I do think, so kind of like, maybe step back again, kind of give you the relevance perspective. So for context again for your listeners, we've been around for five years now. We've raised $42 million USD to date. I saw that in the brief. That's amazing. Yeah, thank you. Did you get a big vow? We did. We did pretty well, yeah.
Speaker 1Yeah, you're allowed to disclose a vow. Raised $42 million on for what sort of vow?
Speaker 2It must be large. Yeah, it's a good valuation. We're very lucky. And then who are your investors? More importantly, yeah, we have some amazing investors. Bessemer, Bessemer Venture Partners, Inside Partners, Peak 15, which used to be Sequoia, Asia, and then King River Capital here out of Sydney. We've been very lucky to have an amazing set of investors backing us. And so the reason why they back us is over the last few years, we've been very much gung-ho on this. Like in 2023, we announced our Series A with the headline that we believe companies are going to scale with ideas, not headcount. Back then, it kind of sounded like a marketing slogan. Now, it's clear from software engineering that that's happening. And so we've been very opinionated in that direction, and we've been very lucky to work with some of the best companies in the world. And if you check our logo on our website, you'll see some of them to help them do this change. So we're coming from a perspective of actually seeing this work in reality. And it was much harder to do it six months ago than it is now. So I can only imagine how much easier it is to achieve the same outcome six months from now. And what are those outcomes? Work that can be done by machines that previously could only be done by people. And I think that's a net positive thing, by the way, because for every team we've done this, we've seen those teams be able to produce better work at a better quality at a rate that's much more efficient for a business to run. Because I think right now every company is putting a lot more pressure on their people, and I think that's not sustainable. And so actually, I think AI is going to come help solve some of that. But I give you that context because what's my perspective then on how people can roll this out? One of the hypotheses we have when we started the company, the number of ideas, not headcount, was that domain experts are critical. So when we think about AI agents, you should think of them a lot more like employees than you do software. And so what I mean by that, how do we hire and train employees? You don't have an engineer hire a marketing leader and train them how to be a marketing leader. You don't have marketing hire a salesperson and train them how to be a salesperson. You have domain experts who are really good at that role, hiring and training for those roles. So the best way that I believe you can do this in your company is find the people, in your business, who are experts at what you do, who are really, really top performers. And ideally, find the ones that also have a desire for kind of systems thinking, I kind of described it, or curiosity about AI. And just get them to have one goal. Automate more and more of the tasks that they hate doing with AI and set them loose. Make that their number one priority and make them accountable for the performance of their agents or their AI. That's the best way to success because the reality is only you, and your top performers, know how to do your job really, really well. If you just think this is like a technology problem where I'm going to have a bunch of engineers, I'm going to build some product, you're not going to succeed. You've got to treat your agents like employees. You've got to train them like employees. You've got to have someone accountable for them, like a manager of any other team. And so for the business owners out there, find that person in your company. Get them started now. Give them the space to do it. Give them goals. Give them resources. But start right now.
Speaker 1Don't wait. The scary thing about that for a lot of people is that people are going to lose their jobs, of course. You know what I mean? You would have heard, this conversation over and over again. And let's say we've got a BDM team, which says 20. There's three high flyers, 17 okay, which is pretty much normal. You know, you get 80, 20. I mean, it might be six high flyers, but it's usually the 80, 20 rule. 80% of the business gets done by 20% of the people, and the rest, you need them because you need the other 20% of work because you want 100%. But unfortunately, generally speaking, you've got, of that 20 people, you've got 14 of them just sort of average. So, which is okay, but one of the things I just want to ask you now is, I think what you're saying is pick up those six or so high flyers and get them to run the program because they know what to make sure that the AI agent needs to know. Exactly. Because they've been in the game for so long, these individuals, and they're the best performers. What do we do about the other 14 people? What happens? What do you think is going to happen in the world for those people? As AI becomes. A non-negotiable. AI agents continue to become a non-negotiable. Is it a retraining? What do we do for those individuals?
Speaker 2Yeah, that's a good question. Look, I definitely don't have all the answers. I think predicting the future right now is really, really hard. I think the only safe thing for us to predict is that AI is going to become a permanent way of working.
Speaker 1Yeah, so that's a non-negotiable. You can't say, oh, we want to protect everybody, so therefore we're not going to do it because you're going to get left behind and you will end up having nobody.
Speaker 2Yeah. Because all 20 will be out of a job. Exactly. Because you won't have a business. Exactly. So I think that is undeniably true. What happens after that, I think, is there's so many different flavors. Like Jevons Paradox, right? This idea that as soon as you change the supply, it will also impact demand and you'll just all of a sudden get so much more demand. So yes, software is easier to create now, but now we want custom software for every single individual. And so we need to basically make sure that we can meet that demand. So therefore, we still need as many software engineers, potentially, because there's just so much more demand. For this product that we now have on offer.
Speaker 1You're saying with AI, we need more software engineers. Potentially, right? Like I'm just saying, like if we look at Jevons Paradox,
Speaker 2that's what's happening in every other part of technology's shift in the history. Now, I don't know if that's necessarily true here. I think there's probably some differences in AI that apply to that. But what I would say is that AI can make those average performers top performers. Because maybe that average performer isn't a top – maybe it's not that they're an average performer due to a lack of work ethic or a lack of interest. Maybe there's other factors. Maybe it's training. Maybe it's like their experience. Maybe they want to be top performers. They don't know how to get there. Imagine you've now got a set of AI agents that know how to do the job at a really, really high level. And they partner with those people. Now, each one of those people can make sure they're performing those individual tasks at a really, really high level. And then the bits that only they can do, they can focus on being really, really good at that. So in fact, that actually might make those employees more valuable. You might want to pay them more money because now that they're actually delivering more output and outcome for your business.
Speaker 1And the AI agents aren't very expensive either because they're not basically – They basically cost you nothing. Yeah, they're much cheaper. They're cheaper, yeah. And they're only getting cheaper. Well, they're cheaper because they might cost you something in the beginning, but they cost you nothing over time. You're not giving them a pay rise all the time. As they get better and better and better, they're not getting bonuses either. Because it's funny, recently, one of my businesses which we lend money, we've got quite a lot of mortgage – physical mortgage brokers around the world. Around Australia, I should say. And like a lot. And the point you just made was a very important point. We've created an AI agent broker. We've created an AI agent broker who's been trained on just about every policy there is that exists in the country. And as people – we haven't sort of actually launched it to the market yet, but it's being launched. It's sort of being trialed. And as it learns what every single one of our brokers does, it becomes one person or one entity that knows everything that everybody knows. But what we're using it for is to help our brokers. So our brokers can talk to it about what they should do. Exactly. To make them better. To make them more efficient. And those brokers that we've spoken to, we haven't gone and spoke – to be honest, we haven't gone and spoke to our best brokers. Our best brokers do it well anyway. We're talking to our not-so-best brokers. And they love the idea of it. And we've actually made it into audio too. So it actually – you can talk to it like I'm talking to you now and it'll talk back to you. And we've given it a female voice and all sorts of stuff. So it's actually quite interesting. We've been working on this for a while now. no one in our industry is in it, but you're right, it's about making each one of the individuals a better person. version of themselves absolutely that's using ai agents to be a better version of in our case
Speaker 2brokers and the thing is we ask a lot of our people we ask a lot of our teams i don't think there's a single company right now that's not actually asking a lot from their people so how can you support them in that journey is what i'm thinking about and i think as well in this future go-to-market teams are even more indispensable yeah like i'm actually on a mission at relevance to not only create the world's best product when it comes to building an ai workforce but also the world's best go-to-market team because i genuinely believe for yourself or for clients both both like i want to have personally the best go-to-market team i want my customers have the best go-to-market team because to your example of the brokers like at the end of the day the person who's coming in to get the loan of course they want the right information of course they want the best information but that's not why they work with a broker like the relationships and the and the kind of the care that people put into things is critical and i think we all trust exactly like we we all want to work with someone that uh cares about us we we trust it looking for our best interest so if ai can help supplement them to make sure that they never make basic mistakes about facts and knowledge great but then they're there for me to support me on my journey and i think in this next wave companies really need to think about how they engage with customers because if product modes maybe start wearing thin gtm modes i think will be really really pronounced what
Speaker 1modes are they again what you said was this what did you call it like
Speaker 2if the product mode is what was the second one you said go-to-market you go to market teams like they might be your most precious resource they're going to be face to face with your customers building those relationships building that trust you can't
Speaker 1under invest in that if i was to ask relevance ai to come do you go into an organization and do an
Speaker 2audit yeah we typically we have a pretty intense engagement with yeah i'm
Speaker 1talking ai audit of course not not a tax audit whatever of course
Speaker 2yeah no we go in and like day one we're like this is not a customer vendor relationship so partners we want to transform your go-to-market organization and we want to transform your go-to-market organization and we want to transform to be an ai first go-to-market organization let's learn about how you sell let's learn how you do things and the reality is we just had a company kick off um last week and we had four customers attend two from the u.s flew over two from australia and the thing they all said about us is they've never experienced the level of care support from a vendor before and that's because we i truly think like our relationships with our customers are going to be so meaningful because unlike software where your tam is actually um quite limited the amount of value and spend that customers can have on you has a pretty finite limit if you look at sas companies generally at our stage the amount of spend that customers are going to have with agents is going to be huge it's going to be much closer to labor spend than software spend and so i want to make sure we're doing everything to invest in those customers to make them successful and grow with
Speaker 1them when you say the amount of spend that they're going to have with agents is going to be close to what they spend with people yeah wow i don't
Speaker 2understand that like uh how do you mean so if that exists today they're going to be a fraction of the spend that people that companies are going to have on ai because ai is going to solve for outcomes like people do it's not going to solve for like making it easier for me to store my data in a database yeah it's actually going to run my business most businesses are going to be heavily reliant on ai to operate in the future and the people that are working in that business will be acting much more like managers than individual contributors so is this sort of
Speaker 1is ai to some extent at a early stage the stage you're talking about for right now is this sort of um going to replace all those philippine admin people that everybody seems to be using we use them um you know wherever whichever country you know nepalese etc i mean is this the next stage of outsourcing admin it's
Speaker 2it's going to have a really big dent on bpos for sure uh you should explain what a bpo is to audience but uh sorry uh business processing um uh offshore organization like what they basically do is as you said they will help you like let's say you're a bank and you need to have invoice processed you know every single day and someone needs to take that pdf and convert it into some input on a software some sort of software system um those companies will be really impacted by agents for sure because right now the last
Speaker 1couple of years we'd be paying someone in australia 120 grand a year to do that but we get it done in uh philippines we actually do it um we operate this way we're paying them 30 grand they're actually very well educated they're you know they're fantastic and uh so we can get four of those people to do what one was the cost of one here in australia now that's been that's been a big had a big impact on our organization in terms of pnl like it's allowed us to save a lot of money but still write the same amount of business um and uh you're gonna you're gonna move that to ai in the next two years yeah so does that mean that uh those those outsourcing businesses because i do it much cheaper than we pay a philippines person it depends on the task
Speaker 2but it will it's undoubtable it's an it's an like it's without a doubt that by some point in the future it'll be much much cheaper than a person doing it and what's the perfect ingredients for an agent to handle your work it's highly repetitive it's a well-known process so you know actually how to do it um and if you have those two pieces of of ingredients you can have an agent do that and as long as the cost basis becomes lower which it's continuously happening like the cost of gpus is going down the performance uh uh the amount of performance and intelligence you have in a model per dollar spent is is improving it's going to get to a place where you can 100 handle those tasks with agents i think that's without question now again i think the important conclusion to not make here is that that means those people and those organizations can't find new opportunities i think there's just going to be more opportunities created somewhere else typically that's what happens um but it's but it's i think we should be i think we should be intellectually honest and definitely be aware of because transitions can be difficult and if we plan for transitions then we can make sure we succeed in them
Speaker 1is relevant say i finding generally speaking widespread adoption adoption of of all this of your prognosis and also the um the treatment for your prognosis of where things are going in 26 27 28 are you getting across the board acceptance
Speaker 2i am this year increasingly like when i speak to kind of cios or ceos i'm increasingly getting more and more um clarity that this is their goal they know to survive they need to adopt ai they don't quite know necessarily how but they know this is happening i think again claude code and what it did for software engineering has been eye-opening for everybody um and again to all the viewers out there like if you haven't seen what it can do just go do some research it is changing software engineering as an industry at that point you no longer have to you know we don't want to debate the reality or the possibility of this it's happening so every company now is going through existential crisis we saw what happened with the public sas companies right there's been a huge sell-off in the last few weeks people are uncertain of where they're going companies like hubspot they exceeded earnings expected earnings yet they've been hit on stock price why is that well it seems like the markets now are not rewarding you for performing based on or even exceeding expectation they're only rewarding you if you're accelerating and that means that you're kind of able to drive ai in your business so i think there's kind of everywhere and so there's a lot of appetite to solve for this and the good thing is it will solve for a lot of these things like i'm actually very confident that a lot of these problems that we do have with scaling with with challenges with burnout with mundane work ai is going to be a huge contributor to helping solve for that and as an end result like again i think you kind of rightly called this out a little bit earlier but what a lot of people don't realize is think about the better quality that you're going to be able to receive the better services you can receive the more you help your brokers be better at their job the better services you're going to be able to service your customers going to receive and
Speaker 1more likely i can recruit more exactly you're going to hear about it i mean you've got to you've got to be the best at technology these days yeah because
Speaker 2the good ones will come to you and your customers are going to come to you because they're going to say working with a broker from this company is better than anybody else because they never make a mistake they know everything every policy they know every
Speaker 1change yeah and no one individual can know that it's just not possible it's not possible and you know what's interesting in our game and it's slightly off the off the topic but it is interesting because ai has allowed us to better make sure that i give you the best deal now i can't possibly give you the best deal because i don't know all the deals i mean there's 30 odd lenders out there and it's and they policies change every week due to appetite credit impairments you know like concentrated risk all sorts of things whereas the only way you can actually know everything is some sort of um you know machine or some sort of agent that is constantly being updated as long as we make make sure it's updated we give it the material updated for it to learn that's the only way you take your duties properly
Speaker 2the only way because now like let's say it took you let's say it takes you two hours to complete a task to do that really thorough investigation and there's no way you can maybe do that in a natural like state of doing business now you can you can delegate to ai it'll go spend the two hours checking thing by line by line every little thing that you need to do again to the viewers out there like don't think chat gpt where you ask a question that returns a response think a system to which you delegate a problem it will work with and interact with a system over two hours to solve that problem and come to you with a solution that's what we're going to and so now that lender oh sorry the the the person seeking the loan as a consumer is going to get the best advice it's going to get the best perspective it's going to get the best offer it's going to get it much more detailed maybe they'll get personal collateral for them maybe that uh person seeking a loan is maybe less educated about how loans work and we really want to educate them and so we're going to create some custom collateral that they understand you know did you know did you know this yeah yeah based on your profile you're probably maybe your first home buyer you don't know probably these things or you're from sydney versus melbourne these are nuanced answers that the AI can guess you might not know. And all of a sudden, every single one of your clients is getting a personalized experience built for them on the fly based on their profile, based on their needs, their questions. And so as a consumer, I'm excited for that world because you're just going to get a better product. And so not only do you help your team perform better, you help your customers have a better experience. Do you think that, is there any chance
Speaker 1at the moment that AI can make errors? Absolutely. How do we sort of manage that part of it? Is that where we keep the BDMs still in the role to oversee everything? Yeah, absolutely. So like,
Speaker 2again, think of the AI agents that work for your team as employees. Employees make mistakes. Humans make mistakes all the time. I am a CMO at a company. I hire a new marketing manager. Their job is to update my landing page. For the first month, I'm probably asking them to show me every update for approval. After a while, I'm like, hey, Mark, you're really good at updating the website. You make no mistakes. Don't come to me for approval anymore. Just give me the details. Keep doing it yourself. You're going to go through a very similar journey with your agents. Our customers do this all the time. Like when you build your agent and you train it, you train it basically with like a one or two page document explaining how to do a job. You connect it to all the tools it needs. And in relevance, we have the option to say, some of these tools can automatically run. Some of these tools require approval. And so you still have a manager. So you give it some rules. Exactly. And so like, hey, before I send out an email, because I don't trust you yet, I need you to seek approval. And so you have humans approve of that. Maybe after a few months, the error rate for this is so low. This employee is really good. It makes a few mistakes. I trust it to do it automatically. Then you switch to order one. So I think, again, like the faster people frame AI and AI agents in particular as the way they would frame about hiring someone, the easier they will have an easier time of both understanding how it works, but also implementing
Speaker 1it. So it's a good way to finish this off, Daniel. And something I got out of, which is quite good for me. Is that I should be thinking about recruiting AI agents, as employees call it. But they're not employees, obviously, but AI agents. I got to think about it as not saying, oh, we're going to have AI in our business. I've got to have AI recruits, AI agent recruits in my business in those areas which AI can manage, but I've got to have recruits in my office.
Speaker 2Or not my office, my world. Yeah. And why is that important? Because it forces you to then build the organization. Yeah. And I think that's a good point. I think that's a good point. Yeah. So what's that? Someone is going to manage that agent. Someone's going to be responsible for training it. Someone's accountable for the goals. Because the thing is- Like I would in a normal recruit. Exactly. Like an agent, you can't hold it accountable because it's obviously not a person. What you can hold accountable is the person managing the agent. And the person managing the agent is then incentivized to actually train it. So a lot of our customers, we have people who started off working with relevance in the company as like an individual person. Very quickly, they start being promoted into like an AI ops manager role. And then they're like, oh, I'm going to be where they'll have a team of people who help them manage agents. And so all of a sudden, you're starting to see what the new workforce looks like. It's not about Bob being able to send out 100 calls. It's about Bob being able to orchestrate a team of agents as if they were humans and train and intervene. I'm going to give you approvals. Hey, this week, we're going to actually do some different stuff. That stuff we did last week is not as good. I'm going to help now train you on a new process. And so if you think about it in that way, you're much more likely to succeed. If you think about it as like a piece of software, that you're now going to buy and install, and it's magically going to solve your problems, that doesn't exist. No magic bullets, no simple bullets here. So think of it in that frame of reference in order to design your organization for the future.
Speaker 1And so that recruit would fit within the architecture of your business, the same as any other recruit. Exactly. It's interesting. Not so long ago, maybe two months ago, I used one of the platforms to look at a bit of land, a property I own around this area. And I said, I want you to go and have a look at this property. And it sort of encouraged me to think about perhaps either selling it or getting an application approved. I think that AI is not scary if you start using it. You've got to use it. Can I just finally, finally, how do you think the AI will interface with us as individuals in our private life, as opposed to what you guys do at Relevance
Speaker 2AI? I think it'll bleed into very similar ways. Like a lot of the, again, mundane work we do in our personal lives, we do it in our personal life. We do it in our personal life. We do it in our we don't enjoy filing your taxes or scheduling, I don't know, booking a dinner or something like that, or changing the booking or booking flights. I just think all of those tasks that previously we all know how we've done, they're going to be done by your AI assistant agent. That agent is going to have access to your systems like you do. You'll give it access to your login details, basically. It'll access those systems and do those things for you. I think that's really exciting. I think it's cool. The only reason it's not exciting is if you see AI as this monster in the closet you don't quite know and understand. If you just see it for what it is, it's a technology that's really powerful in the sense that it can make real-time decisions about what to do next, and you can leverage that capability to do less mundane things yourself. All of a sudden, it's just another piece of tech that we're all going to have in our lives that's going to enrich a lot of our lives. Now, that it's a lot less kind of monster in the closet, and it's a lot more just a piece of technology that's going to change a lot of things about how we live in a very positive way.
Speaker 1You let an AI agent-personal assistant access your URA details?
Speaker 2Everything, yeah, yeah. Do you do it now? Yeah, so the way I, like we, as I said, like we've launched this kind of chat 2.0 where I can connect it to all of my systems, like email and everything, and the more it has access to, the more it can do for me. So, obviously, I'm running it in relevance, which is very much like we have a lot of security and kind of enterprise guardrails, so I trust it. I wouldn't do that with some of these open-source solutions that people come up. I probably wouldn't trust it there, but. Not at this stage, anyway. Yeah, but consumers will get that. Like, whatever's happening in enterprise today, consumers will get the same level of security and services that those people are getting, and you'll be able to connect it to everything, and there'll be laws in place for protecting your data like anything else. There'll be laws for what they can and can't do. I'm sure governments at going to want to have some say in how to regulate, you know, this industry. So, yes, I think embrace it. Be mindful of security. I think always be mindful of security. I want people to be very conscious of, like, not connecting it to your bank account and then having a bad time, but trust your pick-right vendors, use the vendors, and embrace it, and the more you learn about it, I think the more likely you'll be able to succeed with it. As with any change, again, I haven't lived through many of them, but I suspect people who have lived through it will be able to succeed with it. Big technological changes and shifts have found many ways to generate opportunity from them. I think this is no different to those times.
Speaker 1I think you're right. I mean, you're clearly right. I don't mean that in a condescending way. You obviously know a lot more about this topic than I do, but if you don't get on board, you will be left behind. And I think, just finally, does Relevance AI have anything for smaller businesses who may not be able to afford something like you? Something more off the shelf?
Speaker 2So we do have a self-server on the platform, but candidly, we work best with companies over 750 employees. We have enterprise go-to-market teams. That's our sweet spot. That's who we will accelerate the most. There are plenty of other solutions out there that can help you maybe on your journey. Candidly, as I said, you might not even need a product today. You might need a first start educating yourself. So actually just going and learning the state of the market might help you. It might then help you leverage some of those tools in your personal journey. And once you've done that, step back for a second, reflect on the process in your business and look for solutions. But at Relevance, going back to that people problem, I'm on a mission to build the best go-to-market team I possibly can. We have an office in Sydney. For your own business, yes. For our own business. We have an office in Sydney of 70 people. We opened up an office in San Francisco last year, and we're just opening up another one in Barcelona. So my call out to anyone listening to this, if you want to be part of the journey where we're actually deploying this at scale at real-world organizations, want to see it firsthand, and you want to be part of the journey, then you're welcome to do so. If you want to be in an AI-first go-to-market team, then definitely check out relevance.com slash careers. You can see all the open roles. So do I think people are going to be needed? Absolutely. Do I think people are going to be needed in different ways? Absolutely. So if you want to be part of that journey, definitely come join us. And if there are any kind of small business owners that are potentially at that scale where they're starting to see an inflection point and they have an opportunity to maybe take some of those repeatable processes and automate them with agents, then they should definitely check out some of our content. At the very least, look up Relevance.com. Look up what our customers are saying. Because I think if you look at some of these larger companies that are forward-thinking and that are working with us, it might give you ideas for what you can be doing yourself.
Speaker 1So some of those case studies, do they appear on your website? They do, yeah. Okay, that's great. Yeah. Well, Daniel, congratulations, by the way. Fantastic. A, getting a good vowel and raising all that amount of money, which you're now deploying to your business to make it bigger and bigger and bigger. And for a young fellow, someone of your age, which is, it's amazing, and it's great to see an Australian out there doing that. Well done, mate. Thanks so much, Mark. Appreciate it, yeah. Thank you. Thank you.
PodcastTranscript AI