Steve Hind | Lorikeet: The $50m AI Startup Every VC Wanted In.

Speaker A: Fintech Chatter, the conversations fintech leaders actually listen to. Presented by Tier 1 Executive Search for

Speaker B: Fintech hello and welcome to Fintech Chatter. I'm your host Dexter Cousins and The founder of Tier 1 People, the executive search firm that's dedicated to fintech. Yes, today is Steve Hind, the CEO and co founder of Lorikeet. Lorikeet builds AI concierges from for highly regulated industries like fintech, healthcare and energy. Clients include Airwallex, linktree and Eucalyptus. Along with co founder Jamie hall, they've raised over US$50 million. The first Aussie startup since Canva to have all major VC funds in one round. We get into why most AI support tools are built to deflect rather than solve the real problem. Steve also shares with me his experience gained from Silicon Valley working at fintech giants like Stripe and Bridgewater. And he shares how his experiences have changed the way that he thinks about hiring, especially as we enter the AI native era. So let's get into it.

Speaker A: Steve, welcome to your Fintech Chatter debut.

Speaker C: Thanks for having me Dexter. It's a long time listener, first time caller, good to be on mate.

Speaker A: I've been waiting for I don't know what 360 odd episodes for somebody to say that. Steve, um, it is really great to have you on I think for a few reasons we first connected when you were yeah as ah I would class talent thinking about your next move and it's been absolutely amazing to see what you've gone on and done since then. But I guess for the listeners and viewers who don't know Laura Keat and Steve Hind, do you want to maybe share us a little bit about what you do and kind of um, how the idea came about?

Speaker C: Yeah. So Lorikeet is a platform that um, high complexity, high regulation businesses use to build what we call AI concierges. And those concierges engage with their customers across the customer lifecycle. So that could be inbound support, it could be outbound re engagement onboarding, all manner of different use cases and those concierges engage across the channel of the customer's choice. So phone chat, SMS, email, 24 7. Um, this is a kind of rapidly emerging space and the reason we talk about an AI concierge is this is not about sort of AI support bots. Um this is about a new way of providing your customers with a means of using your product and of getting help um, in getting the most out of your product. M. We work uh, extensively with businesses like financial services firms but uh also healthcare firm, others that have needs um that can't be served by your sort of traditional uh, AI solutions because most of those are built for very simple software as a service or e commerce businesses.

Speaker A: So I guess from that, that kind of you're taking it up a few notches around complexity and nuance. Um, that you know, we often say, oh hey, the limitations of AI, how have you, you know, how have you managed to get around that?

Speaker C: Yeah, so, so the origin of us working on this was um, you know, as we were kind of exploring the space a couple of years ago, um, the AI customer support market has always been incredibly crowded. And what we saw was uh, you know, a lot of solutions doing basically the same thing which is taking a customer question, searching in a knowledge base, finding a knowledge base article and then summarizing it back to the customer. And that sort of FAQ first approach really hits a quite low ceiling because all you can really do is tell the customer how to F off and solve the problem themselves. And that's actually not what your customer wants. Um, and as we looked at you know, our early customers we were working with in healthcare and financial services, um, not only is that not what their customer wants, but most of the answers aren't in the FAQs. What we saw the human operators doing as we were kind of embedded with a support team early on is um, pulling data from a lot of sources, taking actions, applying judgment and we said that's what's actually needed to build a useful solution in this space. And so rather than be boxed in by what the off the shelf technology could do, we said let's build our own architecture to make sure these agents can reliably handle these complex multi step resolutions. Um, and that was our wedge into the market. Like you know, partially just driven by um, being very focused on what we thought the end customer needed but also driven by my co founders experience building large language models of Google Brain. We said like we don't have to take architecture off the shelf, we can build fit for purpose. And that's sort of the wedge that's, that's compounded into where we are today.

Speaker A: Yes. So um, obviously there's been a lot of talk about the SUS apocalypse and. Yeah, but I think what, what really is kind of come through from what you shared there is that you're going into every individual client and doing bespoke solution, is that right?

Speaker C: Um, it depends how you define bespoke. Like I think the interesting challenge in this space is how to build a software platform capable of meeting the exact business logic needs of Each of our customers without requiring a ton of customer engineering. And I think we've threaded that needle quite nicely. Always plenty of room to improve. But when I look at market I kind of see our competitors in one of two buckets. So one is they can do anything but they require a ton of custom engineering to get there, which limits them to seven figure contracts. And that's where Sierra sits in the market. The other end of the market. Um, they can be very fast and easy to deploy as software solutions but you hit a very low ceiling. And so if you're a complicated business that's not like the standard business they built for you, sort of run into roadblocks quite quickly. And we think we've kind of um, threaded the needle. Now what that means is the platform is highly configurable and so there is a need to think hard about what are you trying to achieve and how do you map that into configuration. Our for deploy team are fantastic at doing that. We've launched this year an agent we call Coach that works inside our platform to do that work as well. And this is a great way in which AI is sort of changing the equation in SaaS because AI is very good at doing setup and configuration and testing and so that becomes an accelerant to getting the platform to do exactly what you need.

Speaker A: Awesome. Um, so how did the idea for Lorikeet come about and how did you meet your co founder?

Speaker C: Yeah, so um, I raised money uh, sort of off the back of working at Stripe and then at a climate unicorn called Watershed, um, in product roles that involved either managing or working alongside operational teams and sort of seeing some consistent patterns in how ops happen in fast growing companies. Uh, at the same time I was like playing around with the GPT3 API on nights and weekends sort of before ChatGPT launched and said geez, these LLMs are going to have a big impact. Um, definitely not the only person to have thought that but wanted um, to look at how they would have an impact in particular on the operations of high growth businesses. Um, raised money to look at that, uh, um, uh, connected with Jamie and kind of bought him in to co found the business. Uh, Jamie and I have a bunch of friends in common from university and had knew each other a little bit but sort of um, quickly bonded I think from that common context. Uh, he'd been doing LLM research at Google, kind of got sick of them, not really shipping, um, and wanted to get back to being sort of an early builder. And so that was a natural fit for what we wanted to do. I'd say one of the things we learned early on in the first couple of months was that if we wanted to help operations teams, the single best way to help them was actually just help them get the work done because they didn't need better observability, they didn't need better tooling. They just had two much work to do. Um, and so that's what led us naturally into this, okay, how do we handle customer support? And then at the same time, we saw very clearly that the opportunity was clearly broader than just traditionally what you called support. But we felt that you needed to earn the right to go down that path by delivering value where people most needed it early on. Otherwise, you kind of become one of the, like, AI BS artists because you're, you know, spitting them this big grand vision before you've delivered anything. So we were very deliberate about, let's deliver something first. Let's talk about support. And then what we've been very gratified to see is our customers start to push, pull us into the places we always thought there was opportunity, which is, you know, this more proactive or outbound or broader set of use cases.

Speaker A: Yeah, so you. You kind of landed on resolution rather than deflection as being like this core product kind of thesis. How did you get to that point? Was that just through this kind of test and learn of going into customers and kind of really helping them or like, not.

Speaker C: Not really, to be honest, I. We landed on it by just saying, if we were the end customer, what would we want? And it's a remarkably powerful frame, and we still use it for a ton of our decisions. We actually, in some ways think, um, not in some ways, in maybe the core way, think that the customer we're serving is our customer's end customer. And what we believe is if we do the right thing by them and we build a product that does the right thing by them, good things will happen and will flow upstream and our customers will be happy. So even if one of our customers came to us and said, you know, we want to make it really hard for someone to get through to a human to solve their problem, I don't think we would really implement that. Like, we would really be asking, well, is that best for your. For your end customers? On the flip side, that might lead us to say, well, hey, actually it's a better experience for your end customer if we let AI do this and let it handle this decision and let it make this judgment call, because it avoids a handoff and it's capable of doing this. So we'll push in both directions but, but the common thread is what is best for the end user. Uh, and that's what makes the resolution thing so obvious because you know, if I'm texting in to get help with medication that's gone missing or with a bank transfer that's delayed and the um business is celebrating success, when I give up and walk away, it's corrosive to my relationship with them long term. And so that resolution focus is I think the most important thing.

Speaker A: Has that helped you kind of navigate the noisiness of um, the AI space? You know I think back a couple of years ago everything was like a chatgpt wrapper. Um, then we had the um, big push of AI agents and everybody's going to build their own agents which has proved to be a little more kind of complex than promised. Speaking from personal experience, um, has that kind of, I guess focus helped you to really just kind of double down? You've gone from doing your first engineering hire at the end of 2023 to you now what, nearly 90 people, is that right?

Speaker C: We're about 65. I think there's 65 maybe, maybe some interns and stuff at the moment because we're in the, in the summer break in the northern hemisphere. But yeah, that vicinity of people, you've,

Speaker A: you've raised a fairly significant amount as well. So you know, last round, 60 million USD. Yeah, yeah. Um, so I'm kind of curious because look there numbers that in the US you see every day, but you don't see that every day in Australia. So what do you think has been. And if I look as well at the investors, I mean you're the first business since what canva to have all the major funds investing in you in a series A round.

Speaker C: Well look, um, as a starting point I'd say a good amount of it is a reflection that we're in a fantastic market. So put us aside for a second. You know the um, the fit between what LLM technology can do and what you need for customer facing contact is really, really strong. And so um, you know, I think kind of any half decent team is going to have a good shot of raising some money in this market because what investors are going to see is the market is massive. There are no dynamics that make it winner take all or let's say there can't be multiple winners. And so therefore anyone who's sort of credible in that space has a lottery ticket to win the big game and become an enormous company. Um, so some of it's probably not to do with us. It's just um, the good judgment or good fortune to land into a good market. And then to your point about Australia versus not, we uh, have very intentionally from day one thought of Lorikey as a, um, global or kind of Silicon Valley quality company in square quotes that just happens to be headquartered in Sydney in and you know, increasingly as we grow we have people, you know, in New York and London and other parts of the world as well. Um, and I think that has reflected in um, some of our early approach and has probably made us uh, you know, stand uh, out a little in the Australian market. Although, um, you know, as much as we think very highly of that market, that's probably not the comparative set we look to. You know, we have the um, uh, you know, foolhardiness to measure ourselves against the toughest benchmark we can find of kind of the best Silicon Valley companies. We compete head to head with them, we beat them more often than we lose to them in those head to head competitions. Um, and that's very much where we focus when we think about are we doing well and what standards we need to hold ourselves to.

Speaker A: So you're actually dialing in from London. So you know. Yeah, you've just opened a London office, is that right?

Speaker C: Um, we've had team on the ground in London uh, for a year now. Um, we'll be up to eight people there shortly. So yeah, it's grown a reasonable amount and certainly the traction we've had in Europe has been really encouraging. We invested in there much earlier in our growth curve than you would sort of expect, very intentionally and I think that's, that's paid off. London is a massive hub for both financial services and healthcare and so makes a lot of sense for our kind of user focus.

Speaker A: One of the, um, I think kind of challenges, I think that make people, you know, certainly from a talent perspective, I think make people who've been in fintech this last decade probably stand out from others is the kind of regulate the restrictions that they have on innovation. You know, so Silicon Valley famously said move fast and break things. And you know, in fintech you got to move fast and make things. You break them, you kind of tend to end up in jail.

Speaker B: Right.

Speaker A: You got clients who are all in heavily regulated industries. Does that mean that you've kind of had to, you know, put that lens on as well and kind of be careful about the guardrails that you're, you're operating in?

Speaker C: Absolutely. And I think it ends up Being a, um, you know, a differentiator for us, you know, when we're competing with much more horizontal, less focused solutions. Um, you know, our very first production deployment was with Eucalyptus, who are a telehealth company they just acquired by Hims and hers for a billion plus dollars. Um, is under Australian law illegal to say the name of a medication to a person if you've not prescribed them that medication. So if you're doing customer support for Euclidus, someone emails in and says, hey, do you have a zempic in stock? If you say the word azempic and you reply, even if you're saying, no, we don't have asempic, that's a crime. Um, and so we had to be able to comply with that guardrail from day one. So, you know, the bar we started with was make sure the AI is not committing crimes.

Speaker A: Yeah, I was going to say hallucinations are going to be pretty costly in that environment.

Speaker C: Right. And so what that forces you to do is say, what is the right way to orchestrate the technology in the context of these guardrails? When do we want to have much more deterministic approaches to minimize risk? When do we want to take on risk but manage, uh, that risk down as low as we can? And there are lots of different techniques you can use to sort of fit the purpose, but that's the sort of pedigree that led us to build both a lot of guardrails, but also a lot of flexibility about how to use those guardrails. The thing you tend to find is, um, the sorts of regulatory constraints you need to operate under are remarkably similar across different types of businesses as long as those businesses are high regulation. So an easy example that would be, you know, um, in healthcare you want to be very focused on not giving personalized financial, not personalized medical advice to people.

Speaker B: Yeah.

Speaker C: But then if you think about the Financial Conduct Authority in the UK's rules about personalized financial advice, actually remarkably similar. Yeah. So building a system that's capable of doing one gives you a system capable of doing both. And that's why we sort of focus on that high complexity, high regulation, um, set of companies even though they're in different industries.

Speaker A: Yeah. It's kind of recession proof as well. Steve, having recruited through.com crash GFC Covid from an enterprise software perspective, it's been very difficult to get decisions and get kind of investment signed off. The things that always seem to be recession proof are those software solutions that keep the CEO out of jail rather than the One that will get them more customers. Um, so I think it's a, that you've got not only just financial services, but basically any highly regulated environment.

Speaker C: And then we have, we have two energy retailers as customers.

Speaker A: Yeah.

Speaker C: For instance, for exactly that reason, um, we found, um, the, uh, kind of contrarian take that I've been talking about a lot this year that I believe most people will come to accept over time is actually that AI is a more compliant customer, um, facing solution than people. Um, rule adherence and task adherence from AI. You definitely, with the right settings, can get up higher than people. Um, consistency is higher. You can get the AI to emit its reasoning for every decision it makes, which is harder for humans. If you have a policy change, you can propagate that policy change instantly as opposed to kind of going through a change management and retraining process. So I think in some amount of time in the future, hopefully relatively soon, chief compliance officers will actually be saying, why we moving this to AI? As uh, opposed to saying, what risks are we taking by moving this to AI?

Speaker A: Yeah, I mean, it's a really interesting point, Steve, because it's reflected in the hiring that I've seen this last two years where, you know, uh, if I get a client calling saying, hey, we're looking for a head of compliance, I actually don't want to take on the brief because their expectations now are people that are very, very rare. And what they're looking for is somebody who can interpret these million different shades of gray or interpret legislation that doesn't even exist yet for products that are like crypto and these really complex and they're going into multiple jurisdictions and so they kind of push. And I've seen this with the top talent as well as they've become a lot more commercial.

Speaker C: I've definitely seen, um, both compliance and procurement can be a massive accelerator to businesses in this era or a massive handbrake. And I think, um, one way in which the world context is changing the compliance needs to operate under is if you're dragging a business over the coals about what type of advice an AI might give to a customer, and that's slowing the deployment, um, and stopping you getting, uh, as much out of it as you could. While you're doing that, ChatGPT is giving your customers as much financial and medical advice as they can take. I don't think there's anyone who seems particularly focused on that from a compliance perspective. Not OpenAI, not regulators. I, um, think it's kind of flying under the radar and what that means is if your customers want help and advice and your compliance team is stopping them from being able to get it, they're not just going home and saying, well that's our lot, they're just going to chatgpt. And so a lot of banks, healthcare companies are going to lose customer relationships and be disintermediated because of their own timidity. Whereas um, very kind of future forward compliance officers are saying, well I've got to take into account um, how we can set thoughtful guardrails to make sure the things we're really worried about don't happen. But also the commercial upside of having a high quality solution that earns and re earns our customers loyalty, which avoids that disintermediation, um, via these kind of solutions. And I think um, the difference in potential impact depending on how commercial and effective those leaders are, ah, makes a massive difference because the traditional approach of we can afford to spend six to 12 months analyzing the rollout, um, we can afford, you know, six month procurement cycles, um, it just isn't a fit with where the technology is today.

Speaker A: Mentioned earlier, we first connected when um, you were working at Stripe. I remember ending that call with you and just thinking, shit, he's good, he's really good.

Speaker C: Thank you for that.

Speaker A: And it wasn't just the Stripe experience, you know, the Aussie experience that you had, all the experience that you'd had in Silicon Valley, you know, working at Bridgewater bcg, um, but also, you know, I think how you had been able to take all of those experiences and kind of really just got it. I'd love to talk to you a little bit about the background and kind of what drove the move to Silicon Valley, what you learned through that process, um, and how that's benefited you moving into a founder role for the first time as well.

Speaker C: Well, I mean my um, my career, uh, definitely on paper is a little bit random, but um, it's maybe less random in my head. Which was basically that I um, was fortunate to start my career for a couple of years at bcg. I think that's like a great apprenticeship and training. Um, I also got sick of it because I felt I kind of found um, not being able to actually see things through and implement them extremely frustrating. I sort of was felt I was taking on a lot of stress about how these things would go and then I didn't actually get to see it through. Um, and so, you know, went to graduate school to do an MBA in some ways as a way to just think about what was out there. Um, but Before I did that, I was very fortunate to join a Sydney startup called Quiller as their kind of first or second employee and get the tech bug. So in my mind I was already sort of a person that was going to go into tech. But um, by virtue of funding my own mba, um, I got an offer in my second year to join hedge fund. And um, it was very, very remunerative offer. I um, think I also had an offer from Amazon and Bridgewaters was more than 2x financially. And so I said, all right, well um, you've taken on a large amount of debt, maybe you go and work on this and retire it. And it certainly seemed like a very interesting place and an interesting set of problems. So um, you know, took this in some ways detour into um, macro investing. Now, uh, I think the big benefit of it is it was a very rigorous place with extremely high standards where you did m. Very, you had to think very hard and do very high quality work and also be very obsessed with finding out what's true and not obsessed with being right, which means being open to feedback and trying to be as low ego as possible. And that was great training. But I sort of realized after a couple of years there that like, um, you know, that sort of industry doesn't have secular growth behind it. And so the only way that you move up and build a career there, um, is by sort of waiting for people to retire, um, or you know, kind of, you know, elbowing your way through. Whereas the beauty of tech is you have this secular growth and if you get onto the right train and you kind of join the right, or join the right rocket ship, if I mix the metaphor, um, everyone can win, uh, and you can have a rising tide that lifts all boats. And so I looked to get back into tech. Actually a lot of tech firms were not that interested in me because I was like a few years post mba, I've been working in this random investing space. But um, Stripe, thankfully, what, um, that gave me a foot back into the door in tech and then, um, spent some time at Stripe and it was fortunate to have a good run there. Um, went to join a bunch of extract people at Watershed when they were about 80 people and got the experience of that sort of series A, series B stage, um, again. And then that maybe gave me the bug to go even earlier and kind of go down to um, to start my own business. So managed to kind of draw the pieces together. But along the way I think the um, uh, the thing that's compounded there is I think the amount of time you spend on something is much less relevant to your experience than what you can draw out of it. And so if you're very intentional about, well, what am I learning and how are the things I've already learned? Um, how do the things I've already learned kind of fit together, um, with what's needed in a role? You can sort of have these things compound nicely. Um, I do tend to think if I'm giving people advice in their career, like, looking if you knew exactly what you wanted to do, um, and you and m me back when you left bcg, you get a hell of a lot further by going straight there. I think definitely my career has meandered, um, to my detriment, assuming I knew what I wanted. But given that I didn't know what I wanted and that I wanted to maximize learning and developing skills, I've been pretty happy with the way it's played out. But, um, I think that's always the thing people have to confront is like, do you know what you want? If you don't know what you want, how do you maximize your chance of finding out?

Speaker A: Yeah. Going back to that first conversation we had, Steve, it was all of those things that you've just talked about, which is why I was so impressed. And, uh, the breadth of experience, you know, that it wasn't about, hey, I know this industry and this industry and this. It's, you know, how do I go about solving problems? Low ego, right?

Speaker C: Yeah.

Speaker A: And, you know, we kind of talk about, you know, the T bar and being a generalist. So if you're not going to have, you know, critical thinking and a, uh, low ego first principles approach to solving problems, it kind of doesn't really matter what your experience is. Right. And because the tools now are so available to everybody, you know, say we've had this literally leveling of the playing field. I think, you know, one of the things that fascinated me at the time when we spoke was, you know, I'd read a lot of, you know, Ray Dalio and, you know, what they'd done at, um, Bridgewater around their recruitment process and, you know, the kind of baseball card approach to assessing their people. I think absolutely can work if you're clear about what are the values, the traits, the behaviors, and how do you identify them. And it's probably, I think, one of the hardest things for founders to get right now, which, you know, you've got the benefit of experience at. Uh, um, my next question is, has that now filtered in to your hiring process? And when you look for talent as well.

Speaker C: Yeah, it absolutely has. I mean, I think, um, as a starting point, I think startup hiring is, um, when you do it well, it's a mutual trade.

Speaker A: Um, it is the hardest type of hiring you could ever do, man. Like, yeah, this is what founders don't realize, right? You've literally got 10% of the workforce. That is your talent pool. Right. The other 90% on cut out for it. And they don't, they don't actually want it.

Speaker C: And they don't want it. Yeah, exactly, yeah. Um, and you know, the experience at Stripe was interesting. I was there in a period of just massive headcount growth. I did hundreds of interviews, um, on the Stripe side when I was hiring, um, ended up focusing a lot on hiring, um, product management managers. So, you know, just a multiple interviews per day with directors from Google and Amazon and Microsoft. So just seeing sort of what the talent, um, looks like. The interesting thing about Stripe, though, was every person with the perfect CV wanted to work for you. So your job was to filter through them and find the good ones and then hire them. When you go to a startup, that's not the game you're playing. The game you're playing is the people with the impeccable CVs have an impeccable set of choices. And especially at the moment, you know, like an anthropic or an OpenAI is the obvious place for them to go. So what you have to find is, uh, where do you have the ability to get unusually good talent for what you deserve by offering them an unusually good, um, package? And what that means is typically, can you identify people that you think are unusually smart and unusually capable? And then can you offer them a role that you are able to offer because you see how good they are that the world at large won't offer. And that means you're taking risks on people, but they're also taking a risk on you. And so that mutual give and take is really valuable. I kind of learned early on that, um, Viya, to be clear, making mistakes. If you're hiring someone who doesn't need their role at your company to be a success for their career to go in the direction they want to go, it will always be very hard to get the level of commitment and ownership out of them that you need. If you hire someone who's like, this is my ticket to stepping up another level in my career. This is a new scope, I can take on a new role, a new skillset, a new technology, then they need it to work out and you need them to work out and you have strong alignment. And so that's sort of my, um, kind of thesis around startup hiring is like you should explicitly be thinking, what is the win win that comes out of this working and how to identify that? And if you instead say, here's my list of 2000 requirements and I only want to hire someone who ticks all the boxes, you're usually a bit delusional because, well, why do they want to work for you and not anywhere else?

Speaker A: And I think it goes back to what you said around product development, right? Which is putting yourself in the shoes of the customer, saying, okay, well so what, why should they care? And what's in it?

Speaker C: You're offering a product. What is the value prop of your product that makes it attractive to your buyer? And how do you understand your buyer?

Speaker A: You look at the journey that you've been on. 60 odd people. Now I'm sure your business has been through two or three different iterations. Now that's not a slight, that's like, hey, that's the process of growth. Not everybody that you hired in the early days is the person that's going to be with you in the next three to five years. And I think it's really important to hire people for the appropriate stage of the business and have it that, hey, look, you know, we're going to outgrow each other. Let's kind of, you know, enter, uh, this knowing that is going to be the case. And unfortunately I see this time and again, that is, you know, tech companies have scaled, they've adopted the command and control kind of hierarchical structure. To me, you know, the intuitively, you know, applying all of the product disciplines into hiring is the most natural thing to do, right? Like first principles, what's the problem we're trying to solve? Do we even need to hire? What does success look like?

Speaker C: Yeah, yeah.

Speaker A: And then you kind of go through those design thinking problems and human centered design around, okay, what's in it with them? Why should they care?

Speaker C: Right? And um, the thing that has accelerated this massively in the last even six to 12 months is the big increase in model capabilities. Um, so put aside how that affects Lorikeet's product, if you think about how it affects our organization, it means that if you have leaders who rely purely on experience and intuition, half to 80% of that is now completely wrong. So for instance, they might say, well, you know, we need this many people to do this volume of work. Why that's kind of always been the case. And then if you scratch under the surface. Okay, well, what's the task list that makes up that volume of work? Well, actually, it turns out now a very significant portion of it can be done trivially or better with AI. And so maybe we need a lot less people there, but more people in another place where AI has opened up a new opportunity that can be pursued. And so if you don't have leaders who, um, are able to go back and solve from first principles and get into detail, you will get left behind. And one of the things that we have found that I've really emphasized with our leadership team is if you don't have leaders who are actively adopting the frontier AI technology in their work, they cannot lead their teams effectively because they don't know what's possible and they can't challenge their teams. So historically, as a manager, um, you know, you would be able to look at a piece of work or look at a project plan or look at an estimate, and from your own experience of doing the work over years, m, have a sense of whether that's accurate and ask some critical questions and kind of coach the team. Now that the AI has massively changed what's possible and on what timelines. If you are trying to do that based on your experience from doing the work yourself five or ten years ago, it's just completely out of date. And so, um, what leaders need to do and how they need to work has just massively changed. And you have to be really paranoid about that in hiring because some things have been completely changed, but some things are more kind of timeless and universal. And, um, if you don't have leaders who can tell the difference, you'll get yourself into, into a mess really quickly. And I think the biggest example of that is leaders, you know, coming and running a 2021 playbook of like, you know, step one is hire a bunch of people. And these days, very rarely is step one hire a bunch of people.

Speaker A: Yeah, it's interesting that if I look at all of the AI transformations I've been involved in, you know, unfortunate to, to have hired for some of the most successful ones in Oz. But if I look at the common theme, everybody has said, look, this is like 95% of, uh, people transformation and 5% technology, whereas I think with AI now, you got fewer people to try and keep happy, you know, which kind of the other limits the complexity as well.

Speaker C: You also see a lot of, you know, what I would describe as like, role collapse, meaning, like one person being able to cover a, uh, much broader set of functions than they could previously. And so if you've built organizations with like quite thick silos, it's very hard to get the most out of AI because one of the ways in which AI can be very helpful is liberating people to operate with fewer dependencies. So even independent of the number of people you need, because, you know, potentially it opens up new opportunities for you. Just having the ability to have one person control more of the inputs they need to be successful end to end has changed things a lot and can often make, um, kind of legacy org structures quite a big blocker.

Speaker A: Yeah. You know, the other thing I've, I've been hearing a lot as well is the levels of accountability. But I think what AI is doing is it's democratizing information and decision making. Right. That's enabling people who have true talent, who actually have the best interests at heart for the business and the customer to really fly ahead at a pace where, oh, there's a, you know, kind of, they're trying to control their, you know, their patch of turf are uh, very quickly becoming irrelevant and being shown up. Right. And I think that's one of the challenges for leadership right now. Right. Is that it, it hasn't been helped by all the rhetoric about AI is going to replace everybody and there's going to be no jobs. And uh, the reality is we know there's going to be disruption. Right. We kind of got to work together to try and make that a smooth landing. Um, but there's undoubtedly going to be disruption.

Speaker C: It's interesting, um, we have the benefit of quite a young company of being, I think, quite AI native internally, but also not having some of the, um, calcification you can get in a broader org. And so when we work with some of our customers who might be bigger, older companies than us, um, we don't presume that our situation is the same as theirs whatsoever. But the two things we're able to share with them is one, here's what it does look like. If you have a company that's just has almost no barrier to how much they can use AI and can be fully all in and that can create some benchmarking or inspiration. But I think on the other hand, they're trying to be open with them about where we have struggles internally, even for a company like us, um, helping people to metabolize the rapid rate of change and the way in which that requires them to be continually learning new tools, continually challenging their ways of working. And like all of these things, it kind of rebalances what skills and inclinations are kind of most valuable because, um, if you are someone that gets uh, very comfortable operating a standard operating procedure over and over and maybe incrementally changing it, um, this is a very disruptive time to be at work. If you're someone that enjoys trying new things, testing new things, is willing to burn half a day going down the wrong path in order to get to where it might be five times more efficient eventually. There's never been a better time. But there's different types of people and people can always adapt over time. But um, you know, that kind of balance between who's having the most fun and the least fun at work.

Speaker A: Yeah, it does take time, Steve. And I remember, you know, I didn't know recruitment as we're kind of seeing the Internet really impact the, the workplace and everybody start to have a computer on their desk and we had to train people to figure out how to use a mouse, right. And a keyboard and they didn't know what they were and just simple things like how to use email. Now to think that it took five years for, you know, that to actually, you know, come through the, the workplace, look, the technology's got a lot better, but as humans we've not evolved that quickly in a 20 year, uh, period, 25 year period. Right. So I think this is where, you know, the, the kind of, you know, the, the breaks to some extent I think need to be kind of applied as we figure out, hey, how do we manage this transition successfully. But at the same time the experimentation piece, you know, I think is super important. And at the end of the day, you know, I'd say I, uh, look at the things like, you know, what you guys are doing or you know, the, like some of the other solutions. These are typically jobs where you, they're hard to fill. People tend not to want to do the jobs and if they do, there's a high turnover because they tend to be highly repetitive and they want to then eventually go on and they have to do something different even if they don't want to go and be super creative. There's just only so much you can take of doing the same thing day in, day out. And so you know, how it then frees up people to be creative, I think is that, that is the beauty of it.

Speaker C: Yeah, I think this has been, this is very interesting in kind of customer support. Um, being a frontline customer support rep is something like 100% year on year turnover, 50 to 100%. So it's not a role that people, people join saying like, this is what I'LL do for the rest of my career. And it's also a role that often brings out the worst in people in the sense of, you know, if you need to get through 300 emails a day, um, that means in a, in a shift you can only spend five minutes per email. Which means when you come across the 10% that actually take 20 minutes to solve properly, you don't have an incentive or you don't even have the ability to help them. You find that frustrating, the customers find it frustrating. You get turned into almost like a human battery chicken just pecking out templates to send to people because it's the only way to get through things fast enough. Um, and the whole ecosystem doesn't work very well. So it's very much a space where we've seen our customers say, okay, um, as we're using AI to help us scale, we're seeing improvements in the quality of what we deliver to our customers. But our team now can spend more time on the places where they were previously having to rush and get to a better result. And with the time they have free, we can get them to go and do things that are much higher value within the business. And this has been a privilege for us intending to work with very high growth businesses. Um, people are not really implementing uh, Lorikeet and saying how do we go lay off a Hollywood of people? They're implementing Lorikeet and saying how do I take talent that's being very underutilized and go and redeploy it into places where there's more value in the business? And that's kind of exciting I think. You know, um, when we think about broader kind of AI impacts on society, you know, if it's blowing and wiping out a lot of legal work, that's going to be quite interesting to watch because you have people there who spent a long time becoming highly qualified and highly invested in that skill set. But um, customer support is not really like that. And so I think of all the levels of disruption this one feels, um, pretty win win.

Speaker A: Look, I couldn't agree more. And I think if you look at events last week, SpaceX IPO being one, right at some um, point, you know, we, we've kind of taken as a, as a society and a, you know, humanity in general. I think since we uh, landed on the moon, we've kind of taken our foot off the gas and um, yeah, kind of it's been almost participation trophies in terms of innovation, uh, where you know, they've not really moved the needle beyond. So I Think we are seeing as well, you know, that, you know, the, the opportunity to go and focus on bigger problems. And I think this is where the technology we're not now as a species being limited to just labor. You know, if you look at our evolution, right, and how rapid it's been in this last hundred years, it's remarkable. Think where it's going to be in another hundred years with AI, right? And it's, yeah, yeah, it could be, it could be Terminator, but it could be the most amazing society we've ever had. Right. And I think this is the, you know, the kind of optimist in me, you know, thinks that hey, we're at the beginning of something that is, you know, truly, you know, I've lived through the digital, yeah, well, I'd say the computer age, but I think now we are about to uh, we are at the very beginnings of the true digital age. And that means products get redeveloped. The way in which we operate, you know, infrastructure, everything changes.

Speaker C: I'm enormously optimistic about um, what we can do and where we can go. I think the two things that give me a lot of cause for optimism. One is, um, some of the best AI use cases I see us adopting at Warrakeet, um, are attacking non consumption, not attacking existing work. So in other words, um, you've taken our go to market. I think we are showing up now to sales engagements much better researched and prepared than we were before. So it's not that we used to have a team of people doing prep for our sales calls, it's just that we were showing up not very well prepared because we were busy. Now we're showing up much better prepared. So it's a kind of an increase in quality and an attack on consumption. Um, likewise, if you think about what coding agents do in the software development process, we want to hire as many good software engineers as we can find because each of them can now have a much bigger impact on our customers because we can do a lot more than we could before. So again, kind of attacking on consumption. Um, the second thing that gives me a lot of optimism is I think we've always been on a ladder of moving up task abstractions using tools so that we can get closer and closer to the problem we're actually trying to solve. And so two examples, I remember nearly 20 years ago taking introductory econometrics at university. You still learned how to calculate statistical significance with a pen and paper and a standard error table in the back of your textbook. Um, and you did that in introduction. So you understood the foundations. And by and large from then on you could just use software or just use Excel. Um, the same thing will happen now. Like, you know, people will learn to write code, um, you know, in their introduction course, just they understand the concept of writing code and then they'll probably, probably never write it again and never need to write it again. Um, that pattern has never led to us needing fewer people in these spaces. You know, we have a lot more investment bankers today than we did before the spreadsheet. And the reason being these tools are kind of unlocking people to be more productive. As they get more productive, you get the ability to kind of deploy more of them. So I think a lot of the fear, um, is partially a lack of creativity, which, which I think it's like, that's okay. That's what markets will do. The markets will be creative. Each of us individually doesn't need to. Part of the fear, though is the pace of change, which I do think is potentially different with this tooling. And that's where it will be interesting. But you know, at the same time, people can be very adaptive, they're very skilled. We'll see how things go. But, um, all in all, I'm optimistic. I think there can be some amazing stuff that we do. And like you, I feel like we're on the cusp of unlocking like a whole new level of kind of, ah, prosperity.

Speaker A: Um, now we're going to wrap up, but before we do, you're a CEO, you're a founder, ah, you've raised and convinced some really very tough to convince investors to invest an awful lot of money in you. What are your tips to anybody out there who's either a founder, uh, they're looking to raise, they're thinking about being a founder. As to the things that these investors need to hear, need to see, and how do you keep conviction with them.

Speaker C: So probably my most, um, high value advice, just because I think this is different to a lot of what you hear, but it's very true, is early on in the fundraising process, like early in the life of a company, a ton of your success has already been dictated by what you've done previously in your career and your reputation. Um, now if you're 20, that's much less true. It's going to be all based on the product you built and it's probably going to be a very high bar. If you're in your kind of mid-30s, as I was, a lot of it's what's happened in your career already. So actually the best Advice is, if being a founder is something that you want to do in the future, figure out how to be as good as you can at the job you're doing now and build a strong network through it. Because the thing that made it relatively easy for me to raise money early on was actually nothing to do with my pitch. My pitch was like a three page memo. That was it. It was. It was fortunate that when investors went behind my back and found people I'd worked with at Stripe or at Bridgewater or earlier, those people had good things to say about me. That was not something I could control with my pitch. There's no pitch that gets around that. That was, um, I guess being fortunate that I felt like I worked pretty hard and tried to be a good colleague and tried to be obsessed with solving the right problem and doing the right thing by the business in my previous role. So the biggest piece of advice I could give to someone is, if you want to be a founder, be great at what you're doing now. M. Make sure you have people who will go into bat for you, um, without you asking them. And, um, that will make things a lot easier. Now, if you've not done that, then M. Maybe there are other kind of tactics, but I think people massively overestimate the importance of a pitch and massively underestimate the importance of kind of a track record.

Speaker A: Awesome. Steve. It has been fantastic to catch up. I mean, just amazing, right? Like, you know, I think back that first conversation and now it's a wild ride, right? If I'd have gone back and went, hey, Steve, crystal ball, we would. We both would have probably went, yeah.

Speaker C: Nah, yeah, yeah, no, absolutely. I mean, seven years, it's, um. Yeah. What can happen?

Speaker A: Yeah, yeah. Um, before we do finish up, though, uh, we get amazing talent listening to this, this show. If they're really interested in what you've shared about Lorikeet and they'd like to find out about careers with you guys, where's the best. What place for them to head?

Speaker C: Shoot me an email. Um. Steveauriketecx. AI um, my one tip to job hunters is if you send an email introducing yourself that has, you know, three lines of independent thinking about the business, that will probably take you 15 minutes. Um, you'll be in the top 0.1% of cold applicants. So that's always my advice to people. You know, if you can, you can share two to three interesting questions or observations. You're in the top 0.1% of cold applicants.

Speaker A: Well, Steve, thanks so much for joining me.

Speaker C: Man.

Speaker A: I, uh, really appreciate it. I really enjoy.

Speaker C: Likewise. I appreciate it too.

Speaker A: As always, folks, you can connect with me on LinkedIn. If you're new to the show, make sure that you follow us wherever you watch or listening. And if you're coming back, thanks so much for your support because it really does help me in elevating great founders like Steve and great Aussie startups like Laura Keat to a global audience. Until the next episode, keep well.