Ep # 142: Scaling AI-Driven Support in Real Estate SaaS ft. Lindsay Liu (CEO, Super)

Adil Saleh 0:32 Hey, greetings everybody. This is a Hyperengage podcast. Long time coming for a lot of you know, a lot of these conversations that I was so excited to, you know, get on on the Hyperengage podcast. We are now talking about property management. We're talking about real estate and how SAS is making an impact. There are not lots of SaaS platform the past three years that are so much directly, you know, AI powered in the in the real estate segment. I only recall a couple of them, sort of my fingertips. So it was, it was so, you know, it was so interesting to get someone like, Lindsay, who's the founder of, of Super AI. That's, that's a powered front desk, desk, front end desk for property managers. And of course, it's, it's all like automating their workflow. It's all about saving them time. It's all about making the processes more efficient. Thank you very much Lindsay for taking the time. Lindsay Liu 1:31 Thanks so much for having me. Adil Saleh 1:35 Love that. So just to pre record, like, as we discussed, like, you know, I know that it's AI can replace a lot of these front desks, like tier one, tier two support it has for some businesses in some industries, it has completely replaced the human. They barely have human in the loop. So how do you see this vertical support side of thing, and then how you relate it to the real estate when you started Super? Lindsay Liu 1:59 Yeah, I mean, I think, you know, one of the very interesting things about real estate, and talk about kind of a a resilient industry, is that it will be very, very difficult to replace the human aspect of it. I think that's both on the kind of customer relationship side as well as on the tenant experience side of things. And that's just because I think real estate, fundamentally is, it is a physical asset that needs to be managed, right? And so until we have the ability to have robotics go and and, you know, automatically repair toilet and and do all of that, there's always going to be some human element involved. And we're also talking about people's homes, right? We're talking about the thing that for most Americans, is the biggest monthly expense that they have every single month. And so I think, you know, when we think about the opportunities for automation and efficiency in this specific sector, it's not about taking away access to talk to a human. It's actually about enabling more efficiency and then making sure that when you are talking to a human, they actually have the capacity and the bandwidth to have a conversation with you, right? And so I think it's a little bit of a reframe on how we're thinking about building these AI tools. Is, how am I an AI platform that is fully automated, but in service of making sure that I'm taking away the highly repetitive, highly manual workflows from our customers in order to allow them to do the more high value things for the businesses and for their customers as well. So that's a little bit of kind of zooming in the loop. Our exactly, our our POV on it is there will probably need to be a human in the loop on a lot of these conversations. So when we are doing that, how are we making sure that we are doing so in a way that that person understands the context of what they're being brought into, that we also even understand the urgency of that, right? Can this human be brought into the loop later, or do they need to be brought into the loop immediately? So I think those are some of the things that we're thinking through. Adil Saleh 4:10 Gotcha. And there's another challenge, especially from in the construction management, property management, of course, the manufacturing, oil and gas energy segment has been for for the longest time, even with this big AI revolution. One is like, you cannot just take out the human from the loop. Second is, you know how you can make these products and services self serve, you know, I know we can call it as SaaS software service, but it's, it's meant to be self serve in a lot of ways, for a lot of smaller customers, because you cannot be like, you cannot have like human in the loop for all of your customers. You know. So how did you, you know, create that balance while you built this Super back in 2021 I know that it's not that long of a time. It's not like big of time, but I see that you raise the funding as well. So what was the initial stage, product wise, when you thought about solving this problem and having, of course, a human in the loop, and making this agentic workflows and all of this to make the platform self serve, or maybe make the onboarding seem less so, what kind of challenges you had in the beginning and where you guys at? Lindsay Liu 5:17 Yeah. I mean, I think what's interesting about the phase of this technology is it is changing so quickly. You know, even internally, every single you know, every couple of days, there's something. Who out there that were like, Oh, my God, we have to go and relearn this. And I think just, you know, speaking frankly, from a technology perspective, it means that you have to be extremely open to realizing there's things we need to learn how to re architect, there's things that we need to rebuild constantly, right? There's something that you kind of built in a certain way, in what I call, like the old world, right? Of doing things that doesn't really make sense as an architecture in this new world, or doesn't make sense as a way to think about it. And so there is a constant evolution that internally we need to be very growth mindset oriented around, I think when you take that and then you pair that with a customer base that's a little bit less tech savvy as well. That's a lot of change for them. So our approach actually has been we're a customer service platform. Right at the end of the day. We think about good communication, enabling good customer service. So if that's the thing that we're going to invest in having humans do, that's the area. So we're very we are very white glove. We're very concierge with our customers. We want to teach them, right? How does this work? Why did we do it this way? When we go live with our customers, we monitor with them, and we kind of say things like, hey, this conversation, it feels like maybe we should tweak the way that we're training the agent for this, and we're doing that on a customer by customer basis right now. Do I believe that as these tools get more mature and people get more comfortable and understand the concepts of them kind of more broadly, that it will be much more self serve? Yes, I think what we have found for now is that in order to enable success, the more handheld we can be, and the more educational we can be around how these things learn, the more successful those customers will be. Adil Saleh 7:19 Initially, and it's been How long since this product is in the production? Of course, there was some initial development phase and then ideation and validation, all of this. But how long, like, how many years have you been serving these customers, like, like, with finished product? Lindsay Liu 7:35 Yeah. I mean, we, you know, I think we have a very iterative mindset that we've always kind of wanted to embrace. The last startup that I've worked on was founded by Eric Reese, so we literally worked in the lean startup methodology. Everything is about experiments and VPs, right? And so I felt very strongly when we started, that the best way for us to learn what the needs are and what customers needed was to just get something in front of them as quickly as possible. So, you know, we had a initial hypothesis that communication was the big gap in this market. I joke often that if you look at a property managers review site, I you know the one star reviews are almost always going to because of poor communication, right? They never pick up the phone, never answer my emails, don't get back to me ever. That is the state of the industry right now is they are so overwhelmed they can't even do the basics of that. And so we've always had a strong hypothesis that communication was a big problem to solve. Here we started building essentially like, you know, before the boom of these llms, we started building communication centralization tools. So think more like a Zendesk, but, you know, for for verticalized, for an industry, right? And what we realized when these AI products started coming out is, you know, I think initially, there was a healthy sense of skepticism around, you know, what, what are these tools actually able to do? Is it more than just writing? You know, the the generative AI, think that first came out in consumer products was like, frankly, a little underwhelming. We're like, sure we could summarize this, but like, is that that's not in itself a product concept, right? That's a bolt on to the existing product. And so we, we spent some time really thinking about kind of, what are the use cases for this, what's the next generation of this technology that's going to be coming out and realize that unstructured to structured, right? Being able to do that is what the LMS are at, and that leads itself very, very well to communication. And so being able to have a completely unstructured conversation that's dynamic, that feels like it's personalized, that's instantaneous, is a great customer service communication experience. And so we started thinking about in, you know, in 2024, very, very heavily, kind of thinking about, what is the future of communications product in this space that's anchored in the AI native, actually. And so instead of just saying, What can we bolt in to our product? That is, that has elements of that into it. We kind of said, what are the capabilities now that this unlocks? And so that's, that's what we started building really, really heavily. And we commercialized our first version of that at the end of 2024. Adil Saleh 10:26 24 and but of course, before that, it was not like. And the llns were not like that, capable of, like, building specialized agents and, you know, Train for different use cases, because they were, like, so wide and so, you know, general, as you mentioned, like they were not so much personalized. So. Lindsay Liu 10:43 Yeah, you know, I think even a lot of the issues that and the challenges that people run into now, you know, think about it. Context. Windows were smaller, the memory, the processing speed, right? The ability for you to even, like, you know, I don't think people were talking about an orchestration layer. It was just like, you know, the idea of even having an assistant was not something that came in until, maybe, just like, you know, a year, less than a year ago, people started talking about it, experimenting with them, and they were pretty basic. And so I think there's just, it's been a huge learning curve, which is really exciting, you know, I think I, I liken this to, this will be the scale and of the industrial resolution revolution. This is the next industrial revolution. We are fundamentally changing how we think about what people do and where we input tools to either augment or reimagine the things that they do. Adil Saleh 11:45 Absolutely, I'm in excited times. And you know, the capabilities are growing and growing, and some more competition coming when it comes to LLM at scale. Think of, you know, you serving one 50 million people in 10 to 15 years, like property managers, like then the bigger you know question would be like, how you can optimize the cost, the API costs, and all of that. So now they're like other LMS joining in the market, and they're doing it for significant, like, 1/10 of a cost, you know, with the same, you know, with the same, you know, technical debt, and same level of security and scalability and all of that. So I think as you grow bigger as a SaaS, AI power and vertical AI SAS, like, it's, also about, like, optimizing costs at scale, perfect. So now Lindsay also wanted to talk about, like, now you're going about like, building specialized agents, I know, like in real estate, they're like, 10 or 15 different kind of line of services. So for each kind of services, there's going to be, like, of course, different use cases from the sports support standpoint. And you know, are you also, you know, thinking about as a founder, to going and tapping the adjacent market? Because a lot of lot of these platforms, they're trying to go Multi Product, or multi license, or, I know that it's not like entirely different products, but they are like pretty much adjacent because, given the capabilities of AI, given all that you can do today and all these big platform like you, you pick any category like any category leader, be it Gong, be it like Salesforce, you know, in the CRM space, they're trying to tap as much of of the market cap as possible, You know, because now they have, like, availability, they can, you know, optimize the band. They can do it really, really fast with all the data that they have investment and all of that. In your case, like for startup, it's not, it's not that easy still, but how do you see it in a longer term view? Lindsay Liu 13:43 Yeah, I mean, look, that's a, that's an exciting, exciting problem to get to that right? I think the stage that we're at is we kind of think there's, there's a whole untapped market in our core, ICP, that has been underserved for a really long time. I think even in real estate, and you look at most of the tooling that's out there is for the multi family and apartment space, right? It's just where there's more institutional money, therefore where the venture funding goes, therefore where people build right. And so when you look at the more fragmented profit you know, the individual owner operators, the small landlords, the single family home managers, they've been left without access to these tools. And by the way, single family homes are 60% of the rentals in the US, and so, you know, when you kind of do the the kind of analysis on it, you know, it just seems kind of crazy, but it's because it's, it's more fragmented, and therefore you have a lot more smaller players than centralized large players right now, on the one hand, I think you look at, kind of the traditional kind of way that startups and venture companies evaluate that it, you know, it makes sense go for the bigger players. You can get, get 10,000 units right away with one customer. The sales cycle for that customer, by the way, though, is going to be years, right? So for a lot of these startups, you are dead before you even start default debt, right? And, like, it's even harder, because you're talking about like, a three year sales cycle to get to a pilot with one of these big companies, and you need a lot of intros, by the way, along the way to be able to even get you into the right rooms. And so we have intentionally been pretty focused. I think we found our ICP, we found that there is this unmet need in. This segment of the kind of SMB side of property management, and that also, by the way, you can have a much more efficient sales cycle because it's a smaller company. You're talking to a decision maker right away, there aren't as many hurdles. They have big needs to find some efficiency within their business, and so they're really excited to find tools that kind of match their specific pain points there. So I would say, as of now, you know, we think that there's plenty of runway within the space that we're focusing on. Obviously, there's a lot of really interesting adjacencies. You know, I got into into this space because I invest in real estate as well. So I've invested in a bunch of different types of asset classes, small, multi family, single family. We've done short term rentals. Now have commercial in the portfolio as well, and and I could see a great application being short term rental. You know, the short term rental customers are probably your most those are the most demanding customers, but they all usually also ask the same 10 questions over and over and over again. And I'll tell you, it's, how do I use the thermostat? How do I turn the remote on on the TV? I'm sure you've done it right. You've been in an Airbnb, and you're like, I don't know which remote. There's five remotes. Which one do I turn the TV on? And they'll call you at 10pm at night with that being an emergency, because the expectation is it's like a hotel, right? And so I do think there are some great applications and extensions, and I think what we will ultimately end up doing is there are going to be and there are customers that we have that also have parts of their portfolio that look a little bit different, right? So we may have a customer that has a short term rental portfolio, rental portfolio they manage as well, and I think that would be the way we'd want to test and build alongside them, as you have a built in customer that's willing to give you feedback. So kind of go and turn and look at that and kind of say, you know, how would this apply for you? Do you think it's useful? What if we set up a separate instance where we trained it a little bit differently for that need? So that's kind of how I think about it. And I know and I know, you know, it's, it's hard sometimes, as a founder, to kind of zoom out on all of that when you're so focused on the work ahead. But absolutely, I think there are a lot of the a lot of adjacent verticals that require a similar data structure. Adil Saleh 17:47 They have similar workflow. Lindsay Liu 17:49 Exactly. And because, I think, of like, even in real estate, like, we have to understand that you have a property, that property might have different units within it, and each of those units will have different people right associated with that. So you know, where, where can you apply that specificity in a way that adds more value to that workflow? And I think that's probably where we'll end up looking. Adil Saleh 18:14 Interesting, interesting. It's all about like, at this moment, I would not say thinking short term, but it's about more about like, going narrow, and, you know, doing it at scale, and, of course, capture as much of address as much of the market as possible. And then think about, like, adjacent workflows that are similar, maybe a lot of this, like, when you get mature with this, vertical agents, these are so powerful. And you mean, we have barely seen and, like, I think intercom has built one the fin, they're doing it at scale. I don't think there's any other, any other product that has gone vertical, agent at scale at this point. Because just been a year that the engineers and technological leaders here, they started thinking about this perfect, so excited time in this vertical like, how you make agents specialize with past different workflows industry, then go narrow, and then White to, you know, to have the maximum market cap. Now talk about your you already spoke about your segment, and I'm so glad to see that you're so much married to it. You You know them inside out. You've invested in them for years, and I've definitely, thank you when I'm New York City next, for sure, but you know, thinking about this segment, what kind of workflows you have post sales to measure the success. I know data has a fair part in it, but of course, since you guys are more hands on, it's more about questions and all but all those standardized, repetitive workflows you can definitely measure and you know, and then based on those, those metrics or data points or patterns or behaviors, you can measure success, and, of course, indicate risk, all of that. So how do you do it internally? Lindsay Liu 20:02 Yeah, I mean, I think there is an element right now where there's, frankly, just, there's a qualitative aspect to it that we want to maintain, because, again, it is really new technology, right? And so there's an adoption curve here. So even if we say, look, that was a really successful conversation, did our customers think that was a successful conversation? Because, like, we're like, the AI did its job there, but they might not. You know, I think there's, there's just a lot of nuance. I. Um, especially when we're dealing with customer facing conversational workflows, where you're talking about tone, you're talking about, well, yeah, okay, the agent flipped over the call after they said representative four times. But did it capture the information about why they were looking for a representative before it was able to write? I think there's, there's so many little details there. So there is absolutely a qualitative aspect, which is one of the reasons we are so hands on with every single customer right now. And the best of that is we need to learn what they're thinking about that. How did that call feel when it got transferred to you? Right? And it was a live conversation. Was that a happy customer? Were we transferring you someone that was already upset? Is there a way we could mitigate that better for you? So there's a lot of these, almost like, I think we live in a world of edge cases, essentially, when you're dealing with live, unstructured conversations, because you just do not know when it comes to phone calls. You do not know what that phone call is going to be about. There's less predictability right in in that component of it, at every single conversation, though we track, we have started to take again, you know, unstructured to structured, right? We are structuring a summary of each conversation, so you get a kind of up to three sentence summary of what it was about, and then there's bullet points for was it resolved? Yes or no, what was the next step, right? So did we need to bring a human in the loop? Was there a next step involved? Yes or no, and then what was the sentiment of that conversation? Was it positive, negative or neutral? And so getting that at scale allows you to now see how many conversations are we resolving without a human needing to be in the loop? How many conversations are we resolving that require a human in the loop? How many of these conversations require a live phone transfer? For instance, versus I call this, I think there's like this concept of a soft transfer now for phones where we can now say, Yes, this does need follow up, but it doesn't need follow up right now, I don't need to interrupt the property managers to work for in this moment to do this thing. That can be, actually, this could be an email that they could follow up with this tenant about. And so we're trying to think through kind of those options as well as, obviously, what was the tone, right in the sentiment of this call? Was it, you know, was it a good call? Was it bad? Why was it bad? Was it was just a, you know, it was an emergency. So the topic was, you know, they were really frustrated already. And so, yeah, we're, we're collecting that. And I think that kind of the next job to be done for us is to really think about, how do you display these insights, how do you give your customers access to more of these insights, and kind of coach them on what they can do with with that data as a result. Adil Saleh 23:17 Amazing, amazing. And you know, you already have, like, a system in place, like CRM, and you know your internal support, like using your own platform for that. And you know CRM wise, like that is going to be a source of true, the system of record, where all of these systems, these customers, communication, engagement, qualitative data, as you mentioned, that you can feed today and with the context and and summary, which is the same the customers can do as well, using Super that can be so powerful, but in a longer time, you're, you're, you know, at the back end of every experience, you already have some sort of standardization in place at scale when you know when you do it like more often, in a sense that you can, you can make this successful organization at scale like you know, you're able to quickly measure success. I know you mentioned that you're pretty much hands on during the implementation, and you're going live with the customer. But oftentimes, you know, from maybe few months to a year from now, with this AI, you will, you'll be able to learn a lot of these, these repetitive experiences, and, you know, some of the exceptional behaviors that you see now to make it at scale. Lindsay Liu 24:31 Yeah, absolutely. I think, you know, from a scaling perspective, there's very exciting things again, these tools are exceptional at, you know, like being an analyst, right? Like taking a massive amount of data, ingesting that, and kind of being able to deliver insights on that. I think there's, there's a lot of really exciting things there. I think we're just it's such an early stage that I think for us, because we're so specific in the use cases that we're solving, it would be a disservice for us to try to be more removed at this stage from that, and try to build for scale that doesn't exist, right? It's like, you'll kind of find the appropriate time when that is required to be able to make sure that that happens at the end of the day, though, there is a subjectivity to the type of work that we're delivering, or, again, kind of like a customer service agent that is pretty subjective. And I think even with. Humans, right? The you talk to a customer service agent today, they think I did all. I did my job. I had these things I needed to do, and I completed all of them. You on the other end are like, Yeah, but you didn't solve my problem. And so I think those are the things that we're trying to make sure we're not missing any of those cues, and we're really bringing that like empathy and emotional intelligence into how we're building the product. Adil Saleh 25:56 Now, that's what we say a lot, like, you know, nailing it first and then scaling. You gotta make sure you know them enough to scale on them. So, yeah, interesting. You know this, these notions that you have, and I know, you know, if you make that buy additional top and learn to say no on the top, it's Bottom Line. Line follows throughout. And here's your team, by the way, you know, I see they're like, around 10 associated people on LinkedIn. Lindsay Liu 26:23 Yeah, yeah, we've got a pretty we've been running a really lean team, and that's also been really intentional. I think this is also the age what people are saying you can build a billion dollar startup with one person that feels like a little bit of a stretch to me, you know, I'd like to interview that person and really kind of brought them about that Adil Saleh 26:40 You're talking about alias from, from Biff. Thanks, from Biff. Lindsay Liu 26:45 I think so there's, like, some, you know, there's somebody on LinkedIn or Twitter kind of being like, it's possible to do that, but I do believe you can, you know, I think you can have an amazing product these days. And I think actually, one of the things that we've, we've been very intentional about, is the size of your engineering team, for instance, is not actually an indicator of the speed and success that you're going to be able to work on. Obviously, there are, there are benefits to kind of scaling those capabilities, but to move at the speed, we need very, very specific people that. And you know, as we've been interviewing and hiring for for roles in this new world as I as I call it, I'm not necessarily looking for specific one to one capability matches. As far as like, do you work? You know, we're not even, frankly, it's a little bit less specific to what stack Have you worked in before? What tools have you worked in? It's all about mindset. Are you a curious person? Do you apply creativity to how you do your work? Right? Are you autonomous and able to independently start to kind of design and define kind of new ways of approaching things? If yes, then you'll probably be successful in this new world of bullying, because a lot of the you know, it's like, the reality is, I am not an engineer by trade, but, like, I'm vibe coding, right? And so there's a lot of things that you can now accomplish as a non technical person, which kind of pushes, I think, the technical work a little bit more upstream there as a result, right? And so again, you know the fact that, like, we are all relearning kind of what it means to build in this era means that you kind of want to have a small team that's very aligned on mindset there, and even for the non technical team members, because, again, the other half of the team is, is all customer support and kind of customer onboarding folks, and by the way, like, right now I'm the only full time sales person, right? It's just founder led sales at this moment. So it's all customer support and then engineering, right, supporting the product, even with the team that's non technical, I am pushing them every single day to find, find a way to automate things that are part of your current workflow right now, use these tools, and they're building their own scripts. They're building right like they're building things that they probably would not have been able to empower to be able to do a year ago without access to to these tools that are kind of making their own lives easier, making their customers lives better and moving faster as well. So that's been really exciting to see just how this is something that's affected kind of the entire way we think about team and team structure and how we hire as well. Adil Saleh 29:42 I’m glad you brought this up. Lindsay, you know, and this is something like as founders, we realize it every day, because the technology is changing so fast, capabilities are growing so fast that even we founders that are building the product and solving these problems are trying to keep up with with the technology. Make sure we don't get outdated. Make sure we keep up with the competition and the features and product and customers and all of that while we we hire people and we get to, you know, grow our team. And, you know, we we see for these growth mindset and people, we also see that they're also struggling to, you know, adapt to the pace or capabilities of all of this. And a lot of them are not so much well equipped with the way that you can you need to interact, like prompting, giving them the context, using the AI to, you know, full attention, getting the best responses. Maybe using the best llms or different technologies like build on top of llms, you know, making them more tech savvy. So how much you guys are investing into training and managing because that is super, super important for this new age growth mindset, you know, Gen Z of today. You can call them, like these, these folks, these are really, really smart people, and you can really make them assets in their significantly faster time than ever before. Like, I'm talking about, like, three to four months, you know, you can make you know your assets, and if they're like, as you mentioned, like they're curious, growing mindset, hungry for success, and equally, bought into the problem that you're solving with all the customer centricity. Do you have a gold mine of people, especially in the regions that are less tab like India, Pakistan. Pakistan is the fastest growing by the way. I've got a lot of team members here in Pakistan as well. Romania is growing big in tech, you know, you can get really, really smart, smart people from Pakistan, in these regions that are out in out in Eastern Europe, on a very affection of a cause. But now building a really strong Training Center, you know, and taking some initiatives in the training, have some sort of like documentation or anything. So how you guys are playing it out when it comes to training? Lindsay Liu 31:52 Yeah, yeah. I mean, I look, I think we believe that we have to knowledge share here. I think, I think that this is a time where we actually have a lot of information asymmetry around how to use tools and what tools even exist out there. I feel like I am discovering new AI SaaS products every single day. I'm like, oh, that's kind of cool, right? I'm like, Oh, I'll try that. Or let's learn about kind of how they're doing it. It is, it is kind of non stop. And it's actually, frankly, it's, it can be overwhelming, right, at times. And so I think, you know, it takes a little bit of discipline to say, I'm going to, I'm going to force myself on a regular basis, to go and take a look at what else is out there. Take a look at the interesting ways that people are doing things, going to going to events, going to meetups, right? Talking to people that are doing things differently and and I think it's just, it's a it's a learning moment for everybody. There's just no way that anybody, no matter how long you've been doing this, knows, knows, kind of the de facto ways about going things. And I think that's a embracing that and kind of saying, You know what, I'm day zeroing all of this. That's the mindset that's going to allow the team to grow faster. And I think us as as leaders kind of coming to the table and saying, we don't know all of the answers. We don't know that the way that we're thinking about structuring this is the right way to think about structuring it. But the only way to find out is to do it and to try right. What was that? Adil Saleh 33:25 Yeah, I lost you for a little bit. , Lindsay Liu 33:26 Oh, you did. Okay. Adil Saleh 33:27 Please go on. Yeah. Lindsay Liu 33:28 Okay. You know, and I think the only way to really, to really know is to try, and that's where being very surgical in the MVP, in the kind of thing that you're trying to prove out there, allows you to kind of to kind of Spike whether that approach is going to teach you something that is meaningful. And then you can kind of, you know, figure out how to, how to scale it. Adil Saleh 33:53 Absolutely, yeah, learning from experiences. Experiment fast and learn fast. You know, with technology, it's all about that, you know, you need to make sure you have, like, good experimentation window and lab every single day with different problems that you're throwing at AI. And you have to find it for seed for yourself, you know, and how you solve it, yeah, very interesting. Very interesting. Less. Last thing I wanted to explore with you is, like, you know, getting to see, like, as a founder in the in the real estate segment, in the support, I know that support in general has been, has been a big, biggest competition in Congress open AI, like just so many of these tools. But what makes you excited going to this year? I know we are early in the first quarter for the year 2025, product wise, engineering wise, whether it's team growth and initiative, you're taking investment that you're raising anything like what makes you excited for for sitting this year? Lindsay Liu 34:53 I think there's a couple ways I can answer that. I think the first way is, in the first part of that is, I think you know just where everything is going, as far as what is, what the capabilities are, and other interesting ways that we can unlock those capabilities. So I think, you know, from a product roadmap perspective, there's just like, there's a like, there's a lot of very, very interesting things that a year ago, if you put that on the product roadmap, you would say that's a two year project to get down. Now you put it on the product roadmap, you're like, maybe that's two month project to get done, and that's, like, fascinating, right? Just how you have to think about scoping and road mapping and and I think there's an element of it, which is just because you can also doesn't mean that you should, and so we're trying to be very, very disciplined about that. Where do we kind of shore up our core strengths, versus where do we keep adding capabilities? Because, yeah, it is possible. But we also don't want to fall into the trap of of kind of doing a bunch of things halfway and never really kind of getting to the kind of core thing that those customers want to solve. So I think that's kind of one part of it. I think the second part is that we are seeing that the tooling and the infrastructure side for working with with these llms continues to get more robust. And so that's pretty exciting to see more interoperability right, to be able to kind of have fewer one way doors, essentially in the choices that you're making from a technical perspective and and kind of seeing that that's the push that the industry as a whole is making, as far as standardization, you know, I think this is an opportunity to do it in a, in a in a different way. I'm really excited to see what's going to continue to evolve in how the tools start to get more mature? Adil Saleh 36:56 Yes, absolutely. Like, it's, it's like the capabilities and possibilities are unlimited, like, you can think of robots everywhere on the front desk. You're not even human, and even SAS. You know, a lot of these big players coming with the government funding, with this, for good or bad, with this new administration spending heavily on AI, and maybe slightly different direction, maybe some educational institutes are doing some research. And, you know, a lot of various AI capabilities in the manufacturing, there's so much that can change in everyday life, in general, you know, in the next, like, I would say, like, the next five years, even maybe less. So, yeah, I'm excited time very, very, you know, looking forward to, you know, how it all change, and how we can part, play our part, and make an impact as farmers. Yeah. So thank you very much for your time. It was lovely having you and wish you good luck for all the things that you're so excited and you're equally all of us are excited about, and let's, let's keep following our journey. Lindsay Liu 38:02 Thanks Adil. All right. Have a great one. Talk to you later. Adil Saleh 38:05 You too. Lindsay Liu 38:06 Bye.