Faisal Butt

Faisal Butt is the Founder & Managing Partner of Pi Labs, one of Europe’s leading venture capital firms focused on AI and technology transforming real estate and the built world. Since founding the firm in 2014, he has led investments in ~100 companies across 17+ countries, achieving 20+ exits, and backing founders redefining how real estate and real assets are designed, built and operated.

He is also the Founder of Spire Ventures, his principal investment platform focused on scaling and aggregating real asset–backed operating businesses across sectors including property services, infrastructure and asset management.

With a background spanning venture capital, private equity and entrepreneurship, Faisal has built, scaled and exited businesses while investing across both high-growth technology and traditional real assets. His work sits at the intersection of AI, infrastructure and real estate, with a focus on backing category-defining companies and platforms globally.

He holds an MBA with Distinction from the University of Oxford and a degree in Business Economics and Computer Science from UCLA.

Transcript

0:06

The time for adoption is now, and actually, I've never been more excited about the impact that technology is going to have on the sector. It just happened a little bit slower than I thought it would. But I think you needed a fundamental technology shift. And I believe now is the time. Hi, I'm Susan Freeman. Welcome back to our Property She podcast series brought to you by Mishkonde Raya in association with the London Real Estate Forum, where I get to interview some of the key influencers in the world of real estate and the built environment. Today I'm delighted to welcome Faisal Butt. Faisal is the founder and managing partner of PyLabs, one of Europe's leading venture capital firms focused on AI and technology, transforming real estate and the built world. Since founding the firm in 2014, he's led investments in more than 100 companies across more than 17 countries, achieving 20 plus exits and backing founders, redefining how real estate and real assets are designed, built, and operated. He's also the founder of Spa Ventures, his principal investment platform focused on scaling and aggregating real asset-backed operating business across sectors including property services, infrastructure, and asset management. With a background spanning venture capital, private equity and entrepreneurship, Faisal has built scaled and exited businesses while investing across both high-growth tech and traditional real assets. His work sits at the intersection of AI, infrastructure and real estate with a focus on backing category-defining companies and platforms globally. He holds an MBA with Distinction from the University of Oxford and a degree in business economics and computer science from UCLA. So now I'm looking forward to talking to Faisal about his career to date and the amazing and incredibly fast-moving world of technology and AI in the built environment. Yeah, it's it and some of those, it's it's been delightful to have you there and others who supported us during the early years. And yeah, it's been over a decade now since I've set up this platform with a big vision to go out and back the future of real estate. I'm looking forward to talking more about that today. Yeah, it'd be really interesting just to talk a little bit about um you know how things have have changed. So before before we talk about that, just a little bit about your your background. And I think it's it's probably fair to say that you really have one foot in real estate and one in venture. And I know you come from a real estate family background and you have um a degree in business economics and computer science from UCLA, but what then brought you to to Oxford for the MBA? Yeah, so it's it's been a really interesting journey. Well, early inspirations, I guess, were so I come from a real estate family, as you say, and I've kind of grown up listening to discussions about leases and contractors and all the property speak around the dinner table. But I ended up in California for university and then lived and worked out there for around a decade. And this is, you could say, during the first tech gold rush when Amazon and a whole bunch of other like tech e-commerce businesses were emerging. So that was very, very inspirational to be on the sidelines to be able to witness and be very close to it in terms of proximity. So that kind of led me on to leaving a safe tech consultant corporate job very early. So I've only had two years of proper, you could say, corporate working experience. And I jumped ship early. I caught the entrepreneurial bug when I was 25, I think. And I set up my first startup, which is one of the first kind of like early iterations of like e-commerce, you know, around the 90s, like AI wasn't quite big. So, but it was all about what you could do with selling stuff online. So I tried, I kind of dabbled in that. And uh then realized that actually I wanted to build something of scale, and I didn't quite have all the business knowledge and know-how, despite having come from a business family. So I thought an MDA would be a good route for me. And uh, I was a very international guy. Like I've grown up, I have three different passports. Um, I've grown up around the world. I'm ethnically Pakistani, I'd say culturally American, also a Belgian citizen, because I lived there as a kid. So being in LA, while I loved it, it felt a little bit provincial for me and I needed to be in a global city. So I think there's only two big global cities in the world that I would have I would see myself at one's New York and the other one's London. And so I kind of needed to make a pick. And I applied to a bunch of business schools and I ended up getting into Oxford on a full scholarship. And it was a really interesting scholarship because it was funded by a Californian entrepreneur called Jeff Skull, who's one of the founders of eBay, and he basically funds five students to go to Oxford every year who are going to build businesses that also have some level of social impact. And uh I was selected as one of those, kind of flew over to Oxford, and uh absolutely loved it. And I I've I've actually lots of lots to say by about my experience there, but I'll pause because I'm sure you want to talk about other things as well. So that was an amazing um opportunity. And I wondered just how did that experience shape the way you you think about business? Do you think it impacted your business thinking? I think it did. I think my time at Oxford kind of made me think that anything was possible and anyone was accessible. I'd say that's one of my key messages that I took away because they brought people to you that you wouldn't just otherwise have access to. And then it made me think I actually I I can access these people. And a very tangible example I'll give you is they have a week at Oxford. I don't know if they're still doing it, but they were doing it when I was there, it was called Silicon Valley Comes to Oxford, where they would invite all these, you know, big Silicon Valley entrepreneurs and venture capitalists to come and spend two days with the MBA students. So we would be in, we would be in lectures with them, but we would also socialize with them. And the people who came my year were the founders of Twitter. So Biz Stone from Twitter, Reed Hoffman from LinkedIn, Elon Musk from SpaceX. In those days, Elon Musk wasn't as big as he was, as he is now. And he was right there in the auditorium speaking to us. And we even went out and had drinks with all these people later. So it just made me think, wow, anything's possible and anyone's accessible. And then very soon afterwards, I met somebody who became an early mentor of mine, um, this guy, James Kahn, who was a Dragon on Dragon's Den, met him at Oxford, and then ended up becoming a business partner of his right after I graduated. So kind of like, I guess the key message was it was an amazing networking opportunity, in addition to, of course, all of the knowledge gain. Sounds absolutely incredible. And I just wondered my experience doing an MBA at London Business School was that there was no interest in the real estate sector. Everybody there regarded it as a complete dinosaur. And I mean, I think there was maybe one person um in the whole school who came from a real estate background. Um, nothing on the curriculum relating to real estate. I mean, was that was that your experience at Oxford? Well, I was lucky in that the year that I did my MBA, Andrew Baum had just moved across from Reading to Oxford, and it was his first year teaching the real estate course. And of course, I was gonna jump onto it because I was interested in the sector, just having come from, but I hadn't done anything in the space. I just came from a real estate family, so I was interested in kind of learning more professionally, and I didn't know that I was gonna go into it. But I think Andrew was also, you could say, an early mentor in ways. Well, I'm glad you've mentioned Andrew because I have on my phone a picture with Andrew. We got our honorary doctorates from the University of the Built Environment um together a few years ago. So we're there with the wonderful outfit. And I think is he involved with Pylabs now? Yeah, so the relationship started with him being a professor of mine at Oxford. I then would kind of pick his brain as I was contemplating the idea of Pylabs. And then later on, when I went ahead and started it, he would use a lot of the Py Labs data in some of the research he was producing at Oxford. Eventually, I ended up hiring him as an advisor at Pilabs on the research side. So it was actually quite gratifying to be able to have a professor you look up to and then eventually, you know, working with him and hiring him. So it's it's he's no longer with us, but of course we have a really good relationship and and uh he's kind of like a long-term friend and collaborator. That's fantastic. So why don't you, I mean, just tell us in a nutshell what the vision was for Pylabs, what Pylabs does, how it started, and where you are now. So I was initially when I first joined James Conn after a summer stint, like summer internship type of stint, he just kind of said, Hey, why don't you come on board and uh join me? And he had set up like his family office. And I asked him, like, okay, what's gonna be my role? And he said, I don't know. We don't really have a role for you, but we like you. And why don't you just shadow me? So I just shadowed him for a year and I was in every meeting that he would go to. And this is around this is 2009, so post-GFC. So I just then started kind of looking around at the real estate sector and uh had a bit of an epiphany thinking, why don't we back real estate, new real estate businesses that are emerging out of GFC? It kind of made sense for me because the financial crisis had its roots in mortgages and real estate. So, and there was a lot of change and churn and movements happening in the industry. So it made sense to me that new ideas, new teams, new platforms would emerge, and there wasn't really anyone set up to be backing them. So that was the first investment thesis. And I did that initially as a division within Hamilton Broadshaw, which is James's firm, which we called Spire Ventures, which I then later spun out and basically made my own investment firm. That was my first investment firm set up to be backing traditional real estate businesses that were just doing things a little bit differently. And through that firm, I've backed an investment management business called 90 North that grew from zero to about a billion of assets under management, which we just exited last year. We're backed property management business, we're backed developers. So we've we've been investing in traditional businesses for some time. Now, that's all been with my personal capital. And what I understood was that, and by doing that, by investing in traditional businesses, I realized how inefficient they were, and I realized what all the technology gaps were. And then a few years later, I had the epiphany to start looking at the technology side of real estate. And my realization really was wow, this is the world's largest and most valuable industry globally, but it's way behind when it comes to tech adoption. Surely there's tech startups that could help automate this industry, make it more efficient, and kind of future-proof it. And I initially dabbled in that thesis with personal capital. So I was the first backer of eMove, which was an early online estate agent, which grew quite substantially, but then didn't actually work out. They couldn't raise their pre-IPO funding, and then they actually, you know, like a lot of these startups, went bust. So that was a big early lesson for me, right? I was also the bac of Hubble, which is the first, you could say, commercial office marketplace, a bit like a Zupla for offices. So I was, I was personally backing a bunch of these new generations of tech businesses for real estate, and very quickly it became the go-to guy for anything real estate tech related. And I got inundated with business plans from around the world. And that's when I realized that the idea of the thesis was much bigger than me and the capital that I had. So I decided to set up Pylabs as a dedicated platform to, you know, just focus exclusively on this one thesis to go out and find technology businesses that are shaping the future of real estate and real assets. So uh you were uh pretty early into that space. I mean, what what was that, 2014? 2014, yeah. So my angel investments in this space happened in 2013 in rapid succession. I backed three companies. I backed the first online estate agent, I backed the first online mortgage broker, Trussell, and I backed the first office space marketplace within maybe nine months. And then very quickly I just got inundated. Like, who's this guy who's backing these tech businesses in real estate? And then in late 2000 summer 2014, it was I think while I was traveling between Mayfair and Shortage, because what would happen is I was in Mayfair in my three-piece suit with the real estate businesses that I was with. I would then have to transform mid-like commuting to shortage and dress like a shortage hipster and then sit at the coffee shops there and meet the the founders who are trying to build some new innovative businesses. So yeah, around 2014, I decided, okay, I've got to set this up as a proper dedicated platform. And uh so I ideated it in mid-2014 and launched it in probably Q1 2015. So I think I was the first investor in the space, at least this side of the pond. Yeah, and and just looking, because my first uh podcast guest when we started at the end of 2018 was was Brendan Wallace and Fifth Wall beef. Of course. And I think Fifth Wall started in 2016, so you know you were definitely ahead of the game. I was very early on the scene. So in the early days when you know people hadn't really thought about prop tech, how hard was it convincing traditional property investors that this was something that they should be investing in? I mean, honestly, I remember knocking on so many doors and saying, hey guys, like this industry needs to change, it's going to change, this technology wave is coming. And I think that my the vision was ahead of the reality and the industry was certainly not ready. So I was I was that crazy guy that was knocking on doors and being shooed away as like, you know, I was a complete outsider as well. I'm not from real estate traditionally, you know, other than being from a real estate family. I'm not a chartered surveyor. I haven't worked at any of the big agents, the Savils, the Night Franks, the CBREs, etc. So I didn't have any credibility in the sector. I was a complete outsider. And I'm also I was also a foreigner. I mean, now I've been living for 15 years. So, you know, I'm I'm a Londoner. So I had so much that wasn't going for me. So, but that's often where the innovation and the and the great ideas come from. You have to be an outsider looking in. It's not going to come from the inside. So, how did you break through? I mean, was there who was the first that big company that you managed to get on board? So I started very granular. I realized that because I didn't have that credibility of coming from one of the big real estate companies, I'd have to just go granular. So my first fund, Pilaps Fund One, which we launched in 2015, was crowdfunded on Cedars. We had to go super granular. So a classmate of mine, two classmates of mine at Oxford were Jeff Flynn and Carlos Silva. And they were, while we were all at Oxford together, they were conceptualizing this idea of an equity crowdfunding platform. So when I when they graduated, they then worked on this Cedars, which went on to be, you know, a pretty well-known business. So I just called up Jeff and I said, Hey, listen, I'm raising a fund and I I, you know, I don't really have the contacts. Can I put my fund idea on your platform and see if people might want to partake? And he said, Well, we're not set up for funds. We're set for startups. But he was a lawyer and he he was able to, I said, Well, what why don't you just look at it and come back to me? He just came back in a week and said, I've looked at the legal structuring and I think we can make it happen. So we then were the first fund that went onto Cedars and raised a tiny $1 million fund, which then made 10 investments. And um, people who backed us were literally like friends, family, people we knew, professionals from the industry, like some, you know, I often bump into people and they tell me, hey, I was in that first fund that you raised. And I wouldn't even know them often because they came through Cedars as like just the crowd. But a lot of people from the industry participated. And um, that fund backed companies like Land Tech, which was like an early winner in the space, which we exited for 75X, which is a real outlier return that we delivered in 2023. And we backed a company called Office RD, which we exited for 60x. So we did some really home run type investments in our first fund. And that was like the inaugural fund that got us up and running. So everyone who invested in that fund in aggregate after all the losses, because of course, you know, some companies don't work out in venture portfolios, but because we had two home runs, the overall fund delivered like an 8x money multiple, which was unheard of, obviously, in real estate circles. And that was the circle that we were moving in. Amazing. So presumably after that first fund, it was easier to attract some of the bigger names. I think we had had we had the credibility then in that we had already made 10 investments. So we then went on to raise a larger fund. So the first fund was a million dollars, the second fund two, which came in 2017, was $10 million. It was still a small fund. And this time the types of parties that were participating were more like small real estate companies, property entrepreneurs who had made some money in real estate and wanted to participate in the sector. I think when things really got, you could say, more institutional, were our third fund. So at that point, we had made about 40 investments in this space, we were, which was more than anyone by an order of magnitude, just because we had been going for longer, right? You needed the the time advantage helped. And we had seen some successes. So by the time we raised our third fund, which was a $90 million fund, it was oversubscribed by like 40%. We had to turn people away at the end. And that's one where we had proper institutional investors and happy to talk about who some of them are. Well, I just sort of having a look, I saw that um GPE and um APG have been investors. So I thought, you know, you've clearly got the message home now. And uh Yeah. After knocking on a lot of doors and facing a lot of rejection, we started to get some big names. So we had Greystar come in from the US, we had APG, the Dutch Pension Fund, we had a bunch of the London Stock Exchange listed REITs, like GPE, Ashora, Helical, um, from Abu Dhabi. We got Aldar from Hong Kong, we got Swire and Sino. That was the fund where the big names started to come in because we had the track record and the credibility at that point. And what what year was that fund? It was interesting because we did the first close like three weeks before the pandemic hit. So just at the beginning of 2020, we did a first close, and GPE and Patricia and Ashura and a few others came in. Then the pandemic hit, and everybody was closed for business. And obviously, no money was raised for a year. And that wasn't just me, that was anyone who was in any kind of fund management business. You just We're not going to raise any money. So and we're all stuck at home. So I guess the message to the team was guys, we've got a little bit of money. Let's do two things. There's only two things we're going to do now while we're all at home stuck at home. We're going to be looking for startups to back and build up an initial portfolio in this fund so that anyone, when the markets open up again, which hopefully they will, there's a bit of an initial portfolio in this fund that people can kind of kick the tires of. And let's just continue to fundraise, even though we know everyone's close for business. So let's just do the calls. Let's just build the relationships. And I've never seen this happen, but and Asians in particular, we were having calls with every time zone in the world while we were on Stuck at Home. And we we would have calls. I remember this is real hustle. My Hong Kong LPs, the origination of them was because I agreed to do a webinar, which was at their time. For London time, it was three in the morning. So I had to go to sleep, wake up, get on a webinar at three in the morning, do the pitch. And then that's how we started to originate some Hong Kong LPs. And all the conversations happened on Zoom, and we closed them remotely. And it was unheard of because obviously, with Asian investors, relationship trust in person is so important, but we kind of had no choice. So we just had to come up with a playbook of how to build the business remotely at that time. Well, that's that's an incredible story and real dedication actually doing a webinar at three in the morning. But it was an incredible time, wasn't it? So, I mean, in the like 12 years since you started, has Proptech been adopted as widely as um you had anticipated? I mean, has it really transformed real estate? Or or do you think it's it's just sort of begun to improve things around the edges? And there's a lot more to do. I think that my vision for the impact that technology is going to have on the sector was ahead of the industry and ahead of the technology. And um I think that the industry's been a lot slower to adopt, and that's partly because the technology wasn't quite there. So the tech products that came out in the last five, six years, you could say were just marginally better than what the industry was using. And when something's marginally better, people don't switch, right? Because the the customer, and this doesn't apply just to real estate, but it's all traditional industries, the customer is generally lazy and switching costs are high. So you need to give them a really compelling reason to switch products. I think what's fundamentally changed now in the last two years since the launch of ChatGPT in this post-GPT era, now these new generations of AI-native companies are better by an order of magnitude. They're 10x better, and the industry is paying attention now. So there's like two, you could say, forces that are coalescing. There's a bottom-up, which is the technology has gotten so much better. And so now what people are seeing is worth trialing and piloting and actually deploying because you could get a 10x improvement. And top-down, what's happened is the boards and leadership teams and owners of these companies are pushing for adoption because they also see the value that AI can bring. So there's a top-down bottom of pressure. I think that the time for adoption is now. And actually, more I've never been more excited about the impact that technology is going to have on the sector. It just happened a little bit slower than I thought it would be. But I think you needed a fundamental technology shift. And I believe now is the time. And it's happening very, very fast, isn't it? It's sort of it seems to just be improving like every few months. Yeah. Yeah. It's uh the face of change is accelerating. Um, every six months we're seeing kind of new trends, new breakthroughs. An example of that would be in the last six months in particular, there's been breakthroughs in voice technology, right? So voice AI now sounds indistinguishable from humans. And now that you have that technology available, there's so many products that can be built on top of it. So we've backed, for example, a company called Telios, which is using voice technology for repairs and maintenance requests. So in the olden days, when things break down, you would call up and say, hi, my I have a leak or have an issue with my boiler. And then you have all the customer service agents who are trying to help you fix that. That can all now be done with voice AI. So that's just one application. But we've backed another company called Verbaflow, which is doing it for the student accommodation sector. So students inquiring into the operator requires a lot of communication on email, chat, voice, telephone. That's all AI-based now as well. So and the interesting thing is that's just the recent breakthrough that which is leading to a kind of flurry of new startups emerging. I think in six months' time there'll be another breakthrough. You'll have another flurry of startups. So I'm quite excited about our new fund, which is specifically focused on AI over the next five years. And I think every six months there'll be new AI breakthroughs, and that's what we'll be investing in. So it seems that it's now all about AI. It is. AI and robotics. Are companies, are property companies actually using AI as part of their investment committee decisions? I mean, is that happening already? Well, I'm certainly hearing people talk about how ICs are gonna be AI enhanced or AI powered. Um, it's it's an area I'm paying attention to, my team is paying attention to. So I guess that the thinking goes something like this. A typical investment committee at a real estate firm has, let's say, a handful of people who are gonna make the most important decision that needs to be made, right? Which is what are we going to invest in? Are we going to invest in this asset or not? There's this kind of thought experiment of, well, what if we had, what if one of our investment committee members was an AI to supplement the humans on the investment committee? And the interesting thing about the AI IC committee member is that it would have read the entire IC memo, including all the appendices, which we rarely get the time to do. It would have read all the previous IC papers and looked at how those investments had performed. And it would have access to all the world's macro information. So all simultaneously. So it's obviously going to be the smartest person in the room. So I don't I don't think we're at a point where the AI is going to start making the investment decisions for us, but I certainly think it's going to be a, you could say, like an intellectual sparring partner for the humans to be able to kind of like bounce off their critical thinking of the asset on. It's absolutely fascinating, isn't it? And I think you've said that the most valuable real estate companies are going to be the ones with the best AI rather than the um than the best assets. And I just wonder if all real estate companies have access to the same good quality AI, how will a property company differentiate itself going forward? I mean, there's going to be a period during which these companies have different adoption levels of AI. So I don't think that they're all going to reach 100% adoption at the same time. So I think over the next five years, there'll be a bit of a race to adopt different technologies. So there'll be AI for investment due diligence, right? So the analysts that make up most real estate investment teams. A lot of that DD will be conducted by AIs, and we're backing some of those companies at Bilabs. There'll be AIs that are there for investment decision making or supporting investment decision making at the investment committee level. There'll be AI for fine-tuning your asset management and helping you identify opportunities that the humans haven't identified, particularly in large portfolios, right? Where asset management can can get quite cumbersome and needs to be quite data-driven. And there's AI for property management, facilities management, all the operational bits of real estate. So I think that it's going to it's not going to be like we're going to go from zero to a hundred. But I do think that the ones that are embracing it across the board will just have an operational advantage. You can get to a bid faster, for example, if you have AI supporting your DD. You can get to a decision faster if AI supporting your decision making. You can identify asset management opportunities faster and squeeze more juice out of your assets if you have AI supporting your asset management and so on and so forth. So many examples. I mean, won't we still see the roles that we see today, the agents, the asset managers, analysts, or is AI going to be enhancing those roles, or is AI going to be replacing those roles? I think it's yet to be seen. I think one sector worth looking at is programming. So and this is not real estate, but we maybe we can draw some parallels. For software developers, they can now write, because they're not having to write code anymore, and this is also a recent breakthrough in the last, you could say, six months, AI has gotten so good at writing software that some of the best software developers in the world have put their hands up and said, I can't do better than this. So the AI can write better code than me. So what software developers are doing now is that they're just prompting. They're basically just giving it their business requirements, like this is what I want you to build. These are my constraints, this is the solution, this is the problem I want you to solve. And then they're iterating it. What that's led to is that software developers can now write 20 times the code that they used to be able to do six just six months ago. So they've become 20 times more productive. That would have led people to believe that software developers are soon going to be redundant because maybe they just need to work a fifth of the time if they're 20 times more productive. But um what the stories I'm hearing from the Valley, from the Bay Area, is that software developers are working more than they've ever worked. Because there's this whole, there's new, I think they're calling them AI vampires. The the software developers are so in awe of how much they can produce that they're working through the night to code because the opportunity cost of sleep has gone up. Because if you stay up, you can produce so much now. So it's I don't know, I so the honest answers, I don't know how that's gonna translate into real estate. So if we if you had a role at re in real estate that became 20 times more productive because AI is doing the heavy lifting, what's that gonna lead to? I'm not sure right now. But we'll have more answers in the next six months or so. The interesting thing is it's not happening slowly, it's happening very fast. It does seem to be happening very fast. And um, you know, for I suppose quite a traditional industry that um you know isn't that accustomed to change, it's it's difficult to cope with. Yeah. I do think the junior analyst roles are probably the first ones that can be AI automated, where you have an AI agent that is part of the investment team. That seems to be a no-brainer. Yeah, well, we'll see, um, we'll see what happens. Now I wanted to ask you about uh construction, because just looking at you, I think you've just um taken in your cohort number 14 at Pylabs, and it includes uh five startups which are using agentic AI cross the construction life cycle. And I believe you've got some research coming out shortly on um AI and construction. And since we do seem to have issues with construction in this country at the moment, I thought that might be quite a you know a good topic to look at and be really interested to know what AI is coming down the road that is going to help speed up construction. Yeah, so we look at the entire value chain of the built world. So it could be technologies for investment asset management, property management, but also design, engineering, construction, and all the related services around it. So yeah, we we thought that we would pay particular attention to construction in uh the last quarter. By the way, we thematically kind of like shift every quarter. We're looking at deep diving to a certain area, and it's often driven by our research. So the research paper that is coming out soon, I think it's released in the next two, three weeks, is called the self-coordinating site. And we're examining all the different ways in which AI and agentic AI in particular, the new wave of AI, is going to be impacting construction and the broader kind of AEC space, architecture, engineering, construction. So, yes, and we we we timed our research paper with the cohort, so it's all aligned with the focus area. For example, like one of the things we've looked at in this cohort is um frontline workers. So the people, the actual blue-collar construction workers who are on site often come from different countries. They speak different languages, they're not known to be very good at reporting and using iPads or laptops. So, what that leads to is you have opaque information on the front lines. So the people sitting in the offices of construction companies have incomplete information because they don't quite know what's happening on site. But um, you also have voice AI, which has recently be, you know, crossed a chasm and and uh had a technology breakthrough. So we've backed a company called Unbane, which is part of this latest cohort, which is using voice to communicate with the frontline workers and really understanding what's happening on site just through a voice conversation and then synthesizing that unstructured data from voice and putting it into the system so the people back in the office have more clarity around the actual site operations. So those are the types of things we're starting to look at. You know, this um construction is one of the least efficient sectors globally. It's it's ridden with cost overruns, project delays, low margins. So I think there's a lot to go after there. Aaron Ross Powell And could PropTech um in fact, is it right to call it PropTech anymore? It's all PropTech seems like from a sort of bygone era. I mean, it's all about AI, isn't it? Aaron Powell I do think that the early uh startups that emerged in the space were very kind of related to maybe buildings themselves. So prop tech became the term. It was like the easy term. But as investors like us started to look at the full value chain and look at the construction technologies for construction and all the AI and robotics that can make construction itself more efficient, then as we started to look at uh engineering firms and what efficiencies can be brought about there, as we started to look at architecture, as we started to look at infrastructure, which is a related category to real estate, but it's not quite real estate. It's it's so in the broader category is called real assets. I think that yeah, you're right, PropTech seems a bit limiting as a term now. And uh, you know, we ourselves often go through kind of branding exercises as to like what do we want to call what we're doing. Yeah, I think AI is um is good. And and just for our listeners, you you and I think I refer to agentic AI. Could you just sort of you know tell us how that differs from previous iterations? Yeah. So AI has obviously gone through many different iterations over the years. And broadly speaking, I would break it down into pre-GPT AI and post-GPT AI. So before 2020, end of I said 2023, before 2023, you had the older forms of AI like machine learning, computer vision, deep learning. That's been around for some time. More recently, since 2023, you have post-GPT AI. That's AI that's LLM-based or harnessing the power of LLMs like ChatGPT and Claude and Gemini. And the first products that emerge in that space were calling themselves co-pilots, right? So they were not making decisions, the AI, but they were assisting humans and making better decisions and being a co-pilot to them. However, in the last, you could say six to nine months, AI has crossed the chasm, and certain products are more agentic in nature. And really what that means is that they have autonomy to make decisions and actually carry out the tasks that humans would. So the example I gave earlier where voice AI is taking calls from humans for repairs and maintenance purposes that humans used to take previously, is agentic, that AI agent is actually doing the work. So agentic AI AI startups are seen less as software, but more as workers. So it's like having an AI worker on your team. And presumably they have to be trained to do this. Or is there is there an element of them training themselves? Oh, yeah, they absolutely there's all kinds of training that happens with the AI, but you may have heard and the best type of training is on real-world data, but sometimes you don't have the real-world data to train the AI on. So you start training AI on AI-generated synthetic data. So that's also a thing. And that is a workaround when you don't have real-world data to train on. So there is AI training on data that's also generated by AI. It seems a bit circular, but that's also happening. Okay. So when you're you're speaking to this AI agent, when you phone up about your your repair, it's actually working things out for itself. Yeah. So that in that instance, and that's a company in our portfolio called Elios. Yes, the AI can take the call with the human, carry out the entire conversation. Sounds indistinguishable. You can't tell that you're speaking to an AI. And it will ask you to go look at your boiler and spot the lights or the number that you know shows up on the boiler and then it'll ask the human to do certain things. And then it'll the AI will log the issue in the CRM in the background, which the human would have done. And it'll also then dispatch the right engineer to the site if an engineer needs to be dispatched, and then of course, track the issue. So it does the kind of full circle job of a somebody that would work at a British gas that's dispatching engineers to your home. Okay. So there are going to be call center jobs that go, I imagine. I think call centers now because there's been a breakthrough in the last six months in voice AI, I think call centers are going to be completely overhauled with voice AI. That's just one of the recent breakthroughs. And in order to repower all this data for, you know, all these this new technology, we're going to need a lot more data centers than we have. And obviously there's a lot of discussion going on at the moment as to you know where where they go. Are you investing in data centers or technology that um is going to improve the efficiency of data centers? Yeah, so I'm quite um bullish on data centers. So on the PyLab side, we're looking at and investing in technologies for data centers. So we've invested in a San Francisco-based firm called FluX. It's an early state startup, but they built an AI that can help data centers run more efficiently and reduce their energy consumption and energy emissions. So I'm investing in the technology for data centers. Through my Spire Ventures platform, I'm investing in data center-related services and investment and development businesses. So currently, you know, um can't reveal too much, but currently looking at, you know, backing a pan European data center developer. So yeah, I'd say in both on both sides, I'm looking at the physical side, which is the actual building of the data centers, and on the technology side, looking at investing in technologies for data centers. I mean, do you think physically, I mean, with the delays that we have with, you know, with planning and the problems of actually getting connected to the grid and and water, are we actually are we going to be able to keep up with the demands of AI, which uh, you know, as we discussed, is is just sort of improving at such, you know, huge speed. Yeah, I mean, in Europe, there's a couple of data center markets. You could say like there are these hubs and uh they're known as Flap D. That's um Frankfurt, London, Amsterdam, Paris, Dublin. And I guess within these cities, there aren't too many opportunities of sites available that have near-term power. But for the few sites that are there, there's a lot of capital available to go ahead and acquire them and then you know take them through the planning, et cetera, and and and build them out. But it is basically a power-constrained market, as you point out. And so it becomes more of a power-led development strategy versus a land-led development strategy. Yes, I mean I wonder whether there will be new technology coming along that makes data centers smaller, because at the moment we seem to be wanting to build these superscalar data centers, which obviously require a lot of land and a lot of um a lot of electricity. Yeah. There's I mean, there's two types of data centers. I mean, there's a few different types, but there's there's the AI training campuses that you could build really anywhere, and they could be out in the countryside. But then there's city center data centers that need to be close to the end user because of latency. So, for example, Netflix will not host its servers in the countryside because they want to be close to the actual viewers, consumers of their technology, and therefore they will only take space into a city center data center. And uh the demand for that is only growing. I mean, will we see like micro data centers that can sort of go into like old bank premises and that sort of thing? Aaron Ross Powell That could be the case. I mean, there's like listen all kinds of talk around innovations and data centers, and you have uh a company out of London that is talk is building technology to launch data centers in space. Elon Musk is talking about launching data centers in space. So I think it'll be a really interesting space to watch. And there'll probably be all forms of data centers, the micro, as you refer to, the large AI training campuses further outside in the countryside. I think you're probably not gonna get away from data centers within a city because of latency. You're gonna want to stay close to your consumers. Maybe that's something we can use our redundant office buildings for. Yeah. Yeah. So, Faisal, you have seen, I mean, hundreds, if not thousands, of prop tech um startup businesses. I mean, what what have you you learned about what actually works versus somebody that does, you know, a great presentation? How can you tell if something actually has legs? Yeah, well, we've learned so much along the way, having been doing this for over a decade. I'd say that so much of this is founder-led. I get really disappointed when I see great ideas that I'd want to back because it's an area or a sector that I want to invest in, but the management team just does not have capability, the communication skills to really bring it to life. What we're looking for is management teams that are balanced. They've got the technical capability to build something that's going to be, you know, a kind of standout tech, but at the same time, the communication skills to be able to hire the best people, to sell the vision to investors, and to be able to like secure the big customers. So many teams just don't have that right mix. So a lot of it is just the people and the right mix of skills on the team. They need to be going after a big market. I'm looking for founders that are building for global and not being too, you could say, local in their thinking. So when somebody's just talking about building a UK-based business or a Germany-based business, or, you know, a Sweden-based business, it's just not exciting and big enough. So I want European founders to be thinking a bit more boldly, like a lot of the American founders. Like so when the founders of Uber and Airbnb set up, they weren't thinking about just one state or one country. They were thinking of building a global business. I am just looking for that type of big bold entrepreneurial thinking. And then, of course, execution skills. And you can tell with execution skills, you can tell how much they've achieved in a fairly short time span. If somebody's been going for like four years and there isn't much to show for it, it's not exciting. And you know, a very basic rule of thumb is like take a look at what they execute in the last four years and gives you an indication of what they'll probably achieve in the next four. But if I see very fast traction in a short space of time, it's a good indicator of execution. And there must have been instances where your instinct was that this was something that was going to work and it just, you know, you were wrong. Yeah, that's right. So in venture, the general rule of thumb, which is called the power law, is that for every 10 investments you make, you need to get it right two out of 10 times. But the two out of 10 need to be home runs. So our fund one is a great, you could say, illustration of that. We made 10 investments, LandTuck did a 75X, Office RD did a 60X. The overall fund was a top performer, even though we had like five write-offs or whatever, and a few like companies that didn't really return much. So in venture, it's perfectly fine for you to not always get it right, and that happens, but I guess the key is to keep just learning and refining and not making the same mistakes again. So we're certainly not chasing ideas that are great that are not backed up by management teams that are rock solid, right? So it's much more management team first. And I think the kind of the um evaluation criteria is changing as well. Like it used to be that you had needed a highly technical team to build an AI business, but now with vibe coding and programming itself becoming automated, that is changing. You could have a very smart business-oriented team build a product relatively fast with just prompting. So we also have to keep up with the times and our and our and our criteria needs to evolve with it. And one thing that um, you know, I think about is you know, you hear that jobs aren't going to go because we're always going to need human skills. And so, you know, humans will be working with with AI, but you come across quite a lot of humans that don't really have very good human skills. And I I wonder whether AI is going to be able to develop these as well and actually in many ways be better than humans. I don't know. I can't quite I feel that you know this whole IQ and EQ debate, whereas IQ used to be really important, and you know, not too long ago, your parents would be telling you, you know, study engineering or study computer science or study law, study the technical skills, because you know, generalists don't have much value in the world. I kind of feel like that paradigm has now tilted. It's it's now the opposite. I think that it's the generalist who can flex his or her instinct, intuition, creativity, entrepreneurialism, people skills, and EQ. That's the professional of the future. Because I don't think technical skills matter because the AI is can do it all, and you just need to be able to access that. It's giving us something to think about, Faisal. And I think it's probably a good time to finish. So thank you so much. That was really, really interesting. Thank you for your time. Thank you, Susan. And uh yeah, it was it was great to come in and talk to you today. Thank you so much, Faisal, for talking us through some of the businesses you're investing in and the direction and exponential speed of travel for AI in our sector. And also for giving us some insights into what the future may hold. So that's it for now. I hope you enjoyed today's conversation. Please join us for the next Property She podcast interview coming soon. The Property She podcast is brought to you by Mishcon DeRaya in association with the London Real Estate Forum and can be found at Mishcon.com/slash PropertyShe along with all our interviews and programme notes. The podcasts are also available to subscribe to on your Apple Podcast app and on Spotify and whatever podcast platform you use. Do continue to subscribe and let us have your feedback and comments and most importantly, suggestions for future guests. And of course, you can continue to follow me on LinkedIn and on Twitter at Property She for a very regular commentary on all things real estate, prop tech, and the built environment. See you again soon.