The text below is a transcript of the audio from Episode 54 of Onward, "Real estate AI is a revolution: The launch of RealAI".
___
Cardiff: Hello, and welcome to Onward the Fundrise podcast, where you'll hear in-depth conversations about the big trends affecting the US and global economies. We are recording this on Tuesday, September 30th, 2025. Before we start today's show, an evergreen reminder that this podcast is not investment advice. It is intended for informational and entertainment purposes.
Only with that, let's get on with the show and what a show it's gonna be. This is the, what the hell does Ben Miller have up his sleeve episode? I'm Cardiff Garcia of Bazaar Audio. I'm joined as always by co-host Ben Miller, CEO of Fundrise. Ben, how are you?
Ben: Cardiff, it's been a minute.
Cardiff: Like a magician, you are something up your sleeve under your hat, hidden in your trench coat, something completely new that you're gonna unveil to me, first of all, but also to the listener.
To the viewer, and I can't wait to see it. Do you want to tease what it is that you're gonna be demonstrating for us today?
Ben: I've been talking to Cardiff about this for...I wanna say at least a year. We've been working on it for more than two years, maybe arguably three or four, depending on when you want to name the inception date and card. If you've never seen it, you don't know what I'm gonna show you.
Cardiff: I will be every bit as bewildered as anybody who comes to see this episode via YouTube or on the podcast. So I'll be learning along with the listener and the viewer, and. Asking probing questions and maybe also some stupid questions just because I'll be so confused or befuddled, but hopefully that'll also help clarify things for the listener.
Ben: So we're gonna do a software demo on a podcast. It's gonna be interesting to see for a mostly audio medium, how we're gonna do it, but card of a professional. So here's the. The goal is to unveil this new thing that Fundrise and I we've been working on for a long time, think it's going to be a really big part of our business thing.
It's really cool and I'm excited to show you Carvin. Carter's never seen it, so he's gonna be the avatar along the way here now.
Cardiff: Has the public ever seen it before? No. So this is it. I would encourage people who usually listen to this episode on the podcast to definitely check out the YouTube version once it's live, but simultaneously, you and I, Ben, will be narrating what it is that we're looking at on the screen as we go, and hopefully communicating that clearly to the audience no matter what.
Ben: Experiment. So we're gonna try to do this this way and hopefully we can communicate and do a good podcast episode even though it's clearly also a visual.
Cardiff: Demonstration on an audio medium, and we're just the guys to do it.
Ben: Okay,
Cardiff: Let's do it.
Ben: So lemme give you the background and then I'll show you the product. So Fundrise, we launched more than a decade ago to democratizing the in real estate. And we started out by building a FinTech platform so that anybody could invest into real estate now at a $10 minimum. So we digitized the entire, what you'd call the front end of real estate.
So that's clearing essentially, and all of the transaction payment processing and record keeping, and. There's a lot to make that possible. And mobile apps, APIs, microservices. We used to basically raise money into a fund and invest in other real estate companies. And then over time we started seeing that there was inefficiency, not just in how money was raised, not just how money was managed, but also how money was invested with the real estate.
And we started vertically integrating with real estate. So we would get rid of the GP or the real estate sponsor who runs the real estate day to day. And we did that about four years ago. And as we got into the real estate and managing the multifamily or build to rent directly, we saw that there was a lot of data and a lot of technology that was pretty unimpressive, I wanna say about the 1990s type technology.
So we started building tech for our own real estate.
Cardiff: Describe the tech bent. What specific kinds of technologies were inefficient or just unimpressive that you wanted to build a better version of?
Ben: Let's say you own a, in our case, thousands of apartment units. Normally what you get are these spreadsheets that tell you, here's the rents from last month called rent rolls, and they have the financials for the last year to date, or for last year, and budgets. And so you get. [00:04:00] All this data rolled up into spreadsheets.
And what we found is we've got into it's, Hey, guess what? At these properties, there's millions of transactions happening. At any moment in time, there's a water bill. Somebody spent 5 cents on running the sink, and they have landscaping bills, and there's all sorts of transactions and operations happening at the property, millions of transactions, and all I'm getting is a spreadsheet.
What if we actually sucked in all this data from all these properties and started using the data to help us actually run the properties?
Cardiff: So, if I'm not mistaken, this was the origin of you're initially getting interested in starting to invest in technologies because you yourself had a need for better technology to invest.
Ben: Yeah, we wanna drive better outcomes for our investors. And we were frustrated by how the technology worked at the real estate. And we started building and we built, when we were building and buying thousands of homes for rent. Guess what? You can't put thousands of homes on a single Excel spreadsheet.
Breaks. You have to start building databases for it. On any home, you're tracking a thousand data [00:05:00] points. When we would purchase a home, we would take probably 500 pictures at the home of every single thing. A picture of the appliance. Picture of the wall. A picture of the roof, so that you have this database so you can track it.
I mean, there's all sorts of stuff. We started building, it just made our execution more efficien. And we became obsessed with real estate data. And we did that for about two years. And then as we got more and more mature in our data engineering, data science part of our capabilities, we started investing in some of these data companies.
In our venture fund. We invested in DBT lab, which is a critical part of data modeling invested in high touch. We it in Databricks. We messed in a lot of data technology. At some point about two years ago, we started saying, you know, we could probably build some kind of real estate data product with this, all of our expertise and learning and data.
And we started building that two years ago thinking that it was gonna take us six months, two years later.
Cardiff: That's the planning fallacy, but if you'd known how long it would take, maybe you wouldn't have started in the first place. So it's a good thing you might say [00:06:00] that you underestimated how long it would take because it got you started.
Ben: The reason it took so long is that we kept expanding the ambition because first it was about gathering all this data and there's so much data in the world around real estate, and you're modeling the data, and then next thing you know, you're like, oh, we could just do this next feature. And so at some point we start looking at AI and we say, oh, we can incorporate AI into this.
So we built what I think is gonna be one of the best, if not the best real estate AI products anybody's ever seen.
Cardiff: Very exciting. I can't wait to see it.
Ben: So I'm gonna show it to you and it's called Real Ai and we actually own real ai.com.
Cardiff: Is there anything there? If I were to go to it right now?
Ben: Yes. So I'll show you to you and I'll show you what it is and I'll actually let you demo it and let's just see if it lives up to the hype.
Cardiff: And while you're setting that up, let me just also comment that what I like about this is that it is an expertise of fund rises that developed organically. You didn't go out searching for it. This was developed because you needed [00:07:00] it. You needed a product like this, and you started developing some of your own stuff in-house, but then that meant that you knew what to look for in other companies and other products that they make, and you could take all of that knowledge and then incorporate it into this new thing that you're making.
I have a fondness for stories like that. You see what I mean? It's an interesting investment. It's a little bit unusual, I think, and all the more interesting for it.
Ben: A lot of people, they see that there's AI and say, what can I do with it? And we came at it from the other side of here we have this thing we want to do. Oh, can AI help it? Oh, actually. Wow. And so I say to people internally, it's like Tetris, we kept playing it and things kept fitting together and we're like, oh, we should just add this next thing.
'cause everything fits together so nicely. So here is the website.
Cardiff: We are there. Now we're looking at it. There's a search bar underneath a sign that says, your real estate AI analyst understand any property or place with unprecedented data and speed, and it's a nice, clean, minimalist homepage is what we're [00:08:00] looking at now.
Ben: And so what we have been building and have an ambition to do is to build a real estate analyst that's an ai. So a lot of the work in real estate is analysis. I would say the majority of real estate work is analysis of different kinds, acquisition analysis and asset management analysis and operations and leasing and things like that.
And that's all very analytical work. And so we built an artificial and intelligent analyst. So I'm gonna log in and I'm gonna show it to you, and then we're gonna give it a test run. See if it can put it through its paces.
Cardiff: You're logging in now, you're there. There's a few new things that have popped up. It says, understand any property or place. I gotta be honest. My immediate instinct is to try to look up some places that I've lived at before, like my childhood home, things like that. I know that's not exactly what this is probably for, but that's the tempting thing.
Ben: Let's put in your childhood home, Cardiff.
Cardiff: Okay. Are you ready for my address? I'm saying this on the airwaves, but that's okay. Ready? Here it is. 1 0 1 1 1 Wood Song Way. That's [00:09:00] W-O-O-D-S-O-N-G way, Tampa, Florida. 3 3 6 1 8 and it already popped up. Great. My family does not live here anymore, by the way. It's been sold and resold a couple of times since then, but there it is.
Wow. It zoomed in, there's a map there, and then there's a bunch of words on the side of the map, but it found the house immediately, and I recognize exactly where that map is in the Carrollwood neighborhood of Tampa, Florida, where I grew up. Okay, so what am I looking at here?
Ben: He wants to know what you wanna do. I know better things might wanna try, but basically you can do all sorts of stuff a real estate analyst could do about. As property. So you could look at property valuation, property investment. You can analyze the neighborhood. You could look at comparable properties.
You look at the rental market, you could rent it, you could buy it, you could lend to it. So there's a lot of things you might do with a real estate property. So real AI is trying to find out, well, what do you want to know? What's the analysis you're trying to do? So you're probably a home owner or home buyer, but what kinds of things would you like to know about this property?
Cardiff: I personally already happen to know that it [00:10:00] is an owned home by people who are not my family because we sold it, I think back in 2009. Since then, it's turned over a couple of times, including by some investment company that I think completely. Changed it up and then sold it for, I won't give you the numbers, but a lot more money than we had originally sold it for back in the day, and this was just a few years ago.
So what I'd be curious about right now is just about how much would this property be worth if it were to come back onto the market and let's say I wanted to buy it to go back to reclaim my childhood home. Like about what would I have to pay for it?
Ben: So I'm gonna ask you property valuation.
Cardiff: That's literally what you're typing here. I'll tell the audience you've typed the words property valuation and show the data that makes you conclude to the value of this house.
Ben: Okay, let's see what it.
Cardiff: It's doing what AI does, which is it thinks it's got the dots coming up, and now we're just waiting for the answer to come up.
It says, retrieving data gathering valuation data. This is great. This is kind of like using chat GPT or one of the other LLMs. [00:11:00] If you're somebody who uses it and you've asked for something that's detailed, oh my God, it's just popped up the answer. Ben, you wanna tell us what it's saying here?
Ben: So it's doing a bunch of work here and as you're saying, it's chugging away. So what's happening here is there's a lot of different. AI subagents running around doing different analyses and pulling that together for you, because any analysis is gonna be based on other supporting analysis. So for something like this, it knows the home.
It says the property is a five bedroom, six bathroom, single family house. Is that approximately right?
Cardiff: Definitely.
Ben: Okay. It was last sold in November, 2023 for 1.55 million, so it knew the house. It's proposing a current value of 1.626 million, a price per square foot of $287 a square foot. It's saying what it things it would rent for, and it tells you what the tax assessment is.
Cardiff: Wow, that's fascinating. That's really fast and it pulled it up quickly. But there's also all kinds of charts and stuff here too. This is impressive.
Ben: So now it's doing the Tampa Block Home Value sales trend. So [00:12:00] it's looking at the home values and you're seeing that there's been a bit of a bubble in home prices during 20 22, 20 23, and it comes back to the same. Underlying trend from before that bubble back in 2020. So this graph is produced by our ai, did a bunch of analysis and produced a graph, and then it editorializes it.
But what you may not know is it's looking at the block you live on and it's giving you the value of the medium home value of just the homes on that block.
Cardiff: It's fascinating too, and I can tell you that this is right because on that block, the size of the homes varies quite a bit, and what it's showing there, by the way, is the. Median home value of houses on that block is about 900,000, which is a good deal, less than the current valuation of my childhood home.
And that makes sense. It was one of the bigger houses on the block. And it also seems to have put all these figures in the context of what has happened to the Tampa housing market over the last few years, which is also very interesting and very [00:13:00] impressive because it, I guess what it's doing is trying to give me a valuation for the house based on how much it might have changed since the last time that it actually changed hands.
So two years ago it was sold for 1.5 million. Now it's valued at 1.6 million, which captures the fact that in Tampa there was that earlier bubble and now it maybe has come down. So the growth isn't quite as aggressive as it was a couple of years ago.
Ben: So the work it does, and any real estate analyst would do is you want to base what your forecast would be on comparable sales. And so went and pulled a bunch of comparable sales and looked at the prices. They tried to find houses that thought were comparable to the house that it's looking at. And then after creating that table with information like the sales price, sales price per square foot, beds and baths, square footage.
It also then writes some editorial about that. And then it went on to look at the neighborhood investment profile. So the Tampa submarket, this submarket, exceptionally affluent and stable. So here's the demographic strengths and market characteristics of the people who live here. So, median household income of people who live here.
Cardiff, [00:14:00] $220,851. Average net worth people who live here, 1.8 million. 75% of people are married, low mobility, they don't move, and 96% of the people have college degree. 83% of the homes are owned or occupied. There's only 17% rentals. There's very little supply growth of new home construction in the last year, and then it does some investment considerations.
But what you're seeing is a pretty decent real estate analysis of your home, looking at the value comps, the neighborhood, and then some investment considerations. In, I dunno if that took 60 seconds, but pretty fast.
Cardiff: Less than that, I think. Very impressive. And by the way, it matches what I can tell you I've seen with my own eyes. 'cause I go back to that block quite a bit to visit some of our former neighbors who I grew up with, like their kids. And I still see them now every now and again who still live there. That area, by the way, it's full of single family homes as is shown here in this analysis.
Not much construction there. As is also shown in this analysis, it's always been an upper middle [00:15:00] income suburb, I guess, and it appears to still be that not much change there in that neighborhood. It's a part of Tampa called Old Carrollwood. So fascinating. Yeah, well done. Very accurate. I like that it shows you all these other homes in the area that are comparable and what they're worth as well, so it gives you a lot of options for where to go next if you want to just check out the area more generally and not just the home.
Ben: We're just getting started here. What we just saw with some pretty basic analysis, we have these modules where we've tried to build out a robust module that does a piece of work. Rather than just asking a question like going off road, you can stay on the road. And so this is something called your house or block, which I just kicked off.
And what it's doing is it's looking. The block you live on. So normally when you think about, as an economist, you know a lot of stuff about a census, you know, maybe an MSA, but normally census is only done every 10 years, and then they update it every year with a sample, a survey sample. So it doesn't get [00:16:00] down to every block in America.
Of course not. And it doesn't do it very often. So what this just did is it clustered. We have a clustering algorithm. In this case, it clustered these homes together. So this is Tampa, so there's a freaking swamp in the middle of it, apparently. So it gathered similar homes together into what is considered a block.
In New York, it would probably be just a city block, but in this case it includes a swamp or some kind of lake or two.
Cardiff: It's a lake. There's a lot of streets that go around that lake as well. This is all very familiar. It's taken me back.
Ben: Yeah, I bet you grew up there. We're not gonna go as slow on this one 'cause there's a lot to show you. But the point is, is that what we've built is something pretty cool, which you can now analyze a city block, and then in this case it's gonna tell you not just the median home of the city blocks and who had the homes there.
In this case it's saying when they're built, apparently 28% of the homes were built in 1970s. 22% were built in 1980s. 20% built 90 sixties. [00:17:00] But then here is something interesting. The average household income of people who live in that block is $298,000. Triple the national median. 75% are married. Their FICO score of the people who live on that city block is 776.
Cardiff: Yeah, so probably pretty good credit.
Ben: Yeah, and 45% have more than a hundred, and their average net worth is 182 million. So think about as an economist being able to know the FICO, the median income and the wealth of every city block in America, updated every month.
Cardiff: I gotta say, I know it's probably not the expressed intent of this website, but for researchers, this could also be a very useful website for the way it combines all this different survey and administrative data. That's fascinating. By the way, it's very well presented, I should say. There's. A lot of charts and graphs showing the trends that we're going through now, comparing when the houses were built, how much they're worth home values versus sales price trends.
It's very visually compelling as well, so [00:18:00] well done.
Ben: I'm gonna read out some of the things, 'cause this is the type of data that real estate people never see and most people don't really get. And we're talk about this at some point. So 27% of the households have a net worth of more than 3 million. 23% have between two to 3 million. You've seen that here. Yet the household counts.
This is across a small number of people, so there's actually less density happening here. The average time a resident lives here is 154 months. Some places people are moving really fast. 22% of people are retired. I'm gonna read this, the resident profile. Your neighbors represent Florida's suburban, elite, affluent professionals and successful retirees who've chosen substance over flash.
These are Mediterranean traveling fine wine drinking sophisticates, who maintain financial advisors, donate to universities and prefer staying in Hilton hotels. Typical resident is a 42-year-old college graduate. 86% have degrees and 10% holding advanced degrees living in a household of nearly three people, including a child on average.
The lifestyle reflects both success and practicality. They're health [00:19:00] conscious enough to choose salads of fries, pod enough to appreciate Greek cuisine, yet busy enough to order takeout regularly.
Cardiff: I definitely don't choose salads over fries,
Ben: You don't live there.
Cardiff: baby. Yeah, this baby doesn't represent exactly my family, but it sure does sound an awful lot like many of my neighbors.
Ben: You haven't lived there in some time. So there's a premium. They own luxury vehicles. They have preferences for pools, high-end goods. They use eBay. Here's the school data. So one of the things people say is, how do you know what people are buying? How do you know what their financial income is? How do you know everything about what they're doing?
So not only do we have household financial data, but we also have household consumer spending data. And we have real estate data. Like here's the schools. Carrollwood Elementary earns a seven outta 10. Here's the property crimes. It's giving you a overall grade B and then a C plus for a property crime.
You have violent crime, property crime, you have the walk score, transit score, bike score. So you start getting really interesting ability to do analysis 'cause you have not just [00:20:00] real estate data, lots and lots of real estate data. You have lots and lots of other kinds of data. And so it starts creating very interesting possibilities that before when you're making an investment or buying a home, figuring out where I wanna live, figuring out where I wanna buy an investment property, it's giving me a lot more kinds of data.
And then of course it can do a lot more analysis 'cause it's got a lot of tech tools and AI that can do very sophisticated kinds of analysis.
Cardiff: This is great. Who do you have in mind as the person who could most benefit from this website? I mentioned that researchers, I think, would be able to use this for their ends, but I'm also imagining some, a prospective home buyer. I'm imagining as well possible investors in a certain part of the country or part of a city even.
What is in the back of your head when you were designing this website as the goal user of the site?
Ben: Part of what happened is that this was very organic. We built it for ourselves and then we also couldn't help, but like when we see an opportunity, we like, well, let's just do that. So we started with institutional real estate investors and then we expanded it and [00:21:00] now we got consumer, a homeowner.
Cardiff: We're at the homepage, by the way, for people following along and now we're back at, yeah, where are we going now?
Ben: Let me stick with the home for a second and then I'll come back to the homeowner. So let's talk with the homeowner. So one of the things I saw is that the home is usually the most expensive investment anybody makes in their life, biggest purchase. And they have a realtor, realtors helping them make a decision, but the realtors typically helping them make a decision about fit and not a financial analysis.
So this is a real estate analyst that's very sophisticated. You'll see it's gonna be able to do institutional real estate analysis, so it could help. Let me do a new one. Show me homes for sale nearby sim similar to this home, and help advise me on what is the best financial investment to make with regards to purchasing.
A home in this neighborhood. So something like, help me both find a home, but also help me make a good financial [00:22:00] decision about the home you're gonna buy.
Cardiff: So it's both asking for data, but also asking for advice on how to use that data. That's interesting.
Ben: Because I found a lot of people really wanna know what's the risk? I'm gonna lose money, or what's the chance I'm gonna make money? Obviously, if you weren't buying a home, you hope to make a lot of money. You're worried about the downside, but that's not the main reason you're buying a home. You're buying a home is gonna fit your family or your lifestyle, and it's hard to have it both ways.
But this thing now can gather a lot of data for you and help you at least make a better financial decision. It's not gonna be better than a home realtor 'cause it's an AI real estate analyst. But you see here it went and found homes nearby that are similar. Boy, you can buy a lot in Tampa. Holy moly, I'd like to buy this house.
God, that's nice. So it made a couple different recommendations of homes. It showed 'em to you, here's the listings of homes you could buy, and then it's gonna go do a market analysis and then it's gonna do an investment metrics comparison. Comparing, basically, here it [00:23:00] is, looking at these five homes price, the monthly mortgage price per square foot, the estimated annual appreciation rate between 3.9 to 5.2%.
And then it said the investment potential and has saying, is it good? Is it excellent? Is it fair? Is it speculative?
Cardiff: Interesting. How does it arrive at those conclusions? What goes into the model's approach to it when it looks at one of these houses and says, I rate this as good, not excellent, but also not bad.
Ben: I went to Google and said, okay, let's go find an apartment building in Tampa. And I found some apartment building, I don't know, somewhere in Tampa. I don't know anything about this place called Legend Oaks. This is where we started, which is, okay, you're gonna analyze apartment buildings and thinking about analyzing an apartment building to invest in.
It's like analyzing a house. It's just way more complicated. So I'm gonna show you how we analyze an apartment building, and it'll help you understand how we'd analyze a house. So we can do different analysis of an apartment building. Here's a property operations review, here's a rental comps analysis, here's a property valuation.
So the point is that [00:24:00] if you're in the real estate business, you know one of the jobs you normally have to do when you're gonna analyze the property is one of these jobs, in this case, rental comps. If you're gonna buy an apartment building or you're gonna operate apartment building, you're definitely gonna want to know what the comps are.
So we built a module that does a rent comps analysis like an analyst would. If you had an analyst on your team with 2, 3, 4, 5 years of experience and you said, Hey, should we buy this property? Go look at the comps. It's gonna do that job that an analyst probably takes them a few hours probably to do a good comps analysis.
And this will do it in, I think it'll take 90 seconds. It could take a little less, a little more, the bigger the analysis is doing, the more data, the more back and forth, the slower it is.
Cardiff: It can grab the data really quickly. It can analyze it very quickly, and then there it is. It took, in this case, less than 20 seconds. I guess. Executive summary, subject property, what are we looking at here?
Ben: Look at it a lot and it's still grinding away. So it knew a lot about the property. It picked these comps nearby on [00:25:00] this map and then put them on a map. Here's s the subject property. It's 416 units we built in 1982, and here are the ones that are staying at are comps. So it has a lot of information about the property.
The size of average unit, it's across 33 buildings. When it was built, when it was renovated even, and it built this comp set. It's building out this table, and now it has the number units, the ears are built, the air is renovated, the average size of the unit, the in place rent, the rent per square foot, and the property occupancy.
And how far it is from the subject. So that rent comms analysis, you can imagine a person that would've taken them a decent amount of time to go get this data and then organize it and present it, and then it did a market overview. Looking at 9,000 units across this sub-market, the zip code. And then looked at the trends, modeled the trends out, and then it did a comps analysis looking at repositioning occupancy performance growth patterns.
So you're seeing that there's a lot of negative rent growth happening in Tampa. 'cause there's been this [00:26:00] oversupply. And then, as I said, we have all this people data, so it's looking at the average household income of people. Who live in these buildings, their FICO scores, the retention rate, the percentage that are below subprime mortgage, FICO scores, and their debt to income ratio.
So this is actually not a high end property is what I'm seeing.
Cardiff: I think it was showing that the median income there was. Something like
Ben: $32,000.
Cardiff: household income average? No, no, no. Household income averaging 23,000 in this area.
Ben: Really, really bad for that. So this is not a wealthy area. It indicates, quote, higher financial risk. So this is like probably workforce housing, class C. So it did what you would think of as an analysis of a property very fast, very thorough, and this would've taken a person. This is replacing some amount of person's time to do this.
Cardiff: A lot of looking things up, a lot of crunching numbers. And by the way, a lot of formatting, [00:27:00] that takes a lot of time as well. And here you just have it all there and it has this map that's interactive so you can see what some of these other apartment buildings or these other investment properties look like just by going there and zooming in and clicking on them as well.
And it's all presented really in quite a user-friendly way.
Ben: It's really cool and what you're seeing, there's so many cool things about this other apartment buildings decided were not comps, but we have every parcel of every property in America database. That's how comprehensive this is. So it's able to pick comps, but it's able to show you, you might say, actually.
Here's this other property, pine tree apartments, and you could say, actually, I think this is a comp you missed that you would to your team. And then you could kick it off and have it start working on that. What I did instead was I kicked off this thing called a tenant intelligence, which is a profile of property residents including financial health and consumer behavior.
So it can go look at the people who live in this building. 'cause one of the things that happens when you're buying an apartment building. It really matters who the tenants are. Ultimately, you have a building, [00:28:00] but the people who live there matter to whether or not it's a good building to buy. And here it's gonna tell you a lot about the people who live there.
It's telling you about their earnings. It's got
Cardiff: Credit scores.
Ben: their age. This is a struggling here. Check this out.
Cardiff: Tell people what this graph shows.
Ben: Our a, I went off and did a massive amount of analysis and built a graph using Python script, I'm pretty sure. And what it is is the income versus he O score for Hillsborough County property. So it's every property in this county and it compares the media income and FCO score of every property.
That is in the every apartment building in this area. So there's some apartment buildings that have high fco, high income, but Hillsborough County doesn't seem like it's got that many. Hillsborough County seems like it's got a lot of lower income. Lower FICO.
Cardiff: It's bun there. I should note by the way, that the graph we're looking at for those listening in podcast format, this is an interactive graph as well, so you can click on the individual dots where they're [00:29:00] plotted to see the specific property and see the median income of people in that property. And the average FICO score as well, you could see the obvious and totally expected relationship as median income goes up, FICO score goes up.
What's really compelling about the chart? In terms of the information it provides is just how many of these properties are concentrated? I would say at less than a $50,000 meet in income, which I think also makes sense for Tampa, which has so many single family homes. So you would expect, I think, more people who are higher earners to be in single family homes as opposed to these potential apartment properties.
Ben: We've added a new dimension to real estate apartment analysis, which is people data analysis of the people who live in or near a property. If you're looking at, you're about two apartment buildings, and they'll say they're similar, similar age, similar everything. One apartment building has high FICO, high median income, and one has low F meaning income.
You'd much rather buy the one that's got the better tenant. Or a lot of [00:30:00] times what happens, especially during COVID, is you got a lot of tenant fraud. So they're saying that their incomes are X, Y, Z, but actually they might not be, and that would also be something you show up in the analysis.
Cardiff: Interesting. By the way, you could use that chart as a way to investigate where there's some fraud going on. If a place says it has very high median income, but the FICO score isn't. Outlier and really, really low, then maybe that's what's going on. I don't know.
Ben: You can also ask it to model these things out. So you might say, show me the impact of FICO scores, which is credit scores on rental rates of every apartment building in Tampa. So now you're asking what is more or less a big data problem. It would be really hard for an analyst to use an Excel spreadsheet and model out the impact of F ICO scores on rental rates on every apartment building in Tampa.
But this thing has this massive amounts of data, massive amount of tooling. You can write software and create a code environment and execute a Python script or some code to solve these problems. And that's something most [00:31:00] real estate people can't do, and they don't also have access to the amount of data we have or the kinds of data we have.
You start seeing how this real AI is a really good real estate analyst and it starts to like, oh man, I can see how this is gonna start affecting people who are in the real estate industry who do this type of work, but do it by hand. Do it manual. So here it is. It's gonna show you the F ICO score and the distribution across Tampa apartment buildings.
It's grinding away. You can imagine that's a decent amount of data that it's gonna grind on. So here it's showing a bunch of analyzed apartment buildings in Tampa, and then it's gonna produce all sorts of analysis around this question. It's doing a lot of analysis.
Cardiff: Yeah, it is. Goes down and down and down. For people listening, there's a ton of graphs, charts. Again, I wanna say that looks like a lot of attention was paid to the user friendliness of this product in the design and that kind of thing. I imagine that was a big part of putting this together because the charts are all very simple, easy to understand.
The tables are as well. I imagine you can probably download some of the source data. Is that doable for some of these?[00:32:00]
Ben: We tried to make it easy for you to be able to take the data, share the data. Lemme just show you some of the stuff's produced. The thing is, is that it can do a lot of different things. So we,
Cardiff: we'll be here for hours and hours and hours. We gotta hit the highlights.
Ben: What's cool here it is looking at FICO scores versus rents in Tampa, and it's picking these properties and then there's this dotted line, the dotted line's called the line of best fit.
And it's saying, here's the regression, here's the relationship between rents and FICO scores. And what you don't wanna do is you don't want to have a high rent and a low FICO score.
Cardiff: Those are the dangerous places.
Ben: You wanna buy things where the rent is low and the FICO site will buy below the line. So it brings a whole new dimension to investing where you can actually look at whole markets.
The way real estate works today is you're looking at points of data. You're not looking at relational databases, so you're not using relational databases in most cases, and then it goes into a bunch of properties and then have different kinds of analysis of the average rental rates by FICO score.
Cardiff: Average monthly.
Ben: So it is a very sophisticated kind of [00:33:00] analysis.
Here's FICO score, which occupancy rates, which actually surprising is that the higher the FICO score, it's slightly negatively correlated to occupancy.
Cardiff: It's interesting. There's a neighborhood comparison it looks like, where it lists things like this chart right here, average FCO scores by neighborhood. All of these neighborhoods are incredibly familiar to me as somebody who grew up in Tampa. But for example, you see that Hyde Park, which has the highest average FICO score in the city.
That's a very famous area in Tampa that's very close to Bayshore Boulevard, which has the nicest real estate in the city, the most traditional real estate in the city. And Hyde Park is the place that has all the nice shops and everything. But it also, I don't wanna say uniquely, but it is unusual for Tampa.
It does actually have quite a few apartment where you can actually live in stuff. It's not just all big sprawling single family homes in that area. It's one of the few, certainly walkable parts of Tampa. With stores and movie theaters and all that stuff in the area, and you could see there, it also has the highest average FICO score.
I'm very impressed by the way that it was able to name and [00:34:00] label the neighborhoods, but the results are not surprising. This is good. This is about what you would expect just based on what I know of the city. So that's really interesting.
Ben: People have gotten skeptical of ai, especially these days. People start thinking AI has gotten maybe overhyped, but what I'm starting to show you is at least for. Our little world, which is real estate. We've built something that most people would be hard pressed to do as good analysis as this thing is doing, let alone is fast, let alone is cheap, let alone is broad.
So it's a good illustration. We can talk about broader implications of like, Hey man, I know people are starting getting cynical about ai, but holy crap, this thing is able to do something that you, Cardiff could have a professional real estate analyst that you could ask questions to for 50 bucks a month.
Normally you're paying $150,000 a year.
Cardiff: And by the way, that's a good cue to address something that we should talk. About which is the general skepticism towards AI is often driven by the worries that it is gonna [00:35:00] hallucinate an answer, that it is going to say something, that it's just repeating from some part of the internet that it's all that's not really accurate, and then it's contaminating the rest of its answer about something.
Do you want to just talk a little bit about the checks that you've put on this model and on this AI to try to defend against that?
Ben: Yeah, so I was gonna get into a little. Side of it because I think it's interesting what it can do or how it's different. Okay. This is gonna a little technical and then you can help me unravel it. So the way that AI works today, some many people ask me, how is this different than chat g bt wildly different.
First thing, chat g BT doesn't have any of this data. Chat g BT is essentially a glorified Google search. We have all sorts of data that they don't have, and it's not just that you have the data, it's well modeled and we'll come back to that. But the way that it works is it does a vector search, so it searches for things that are similar.
It's kinda like Google. It's like a similarity search and a similarity search. It's gonna bring you a similar thing, but if you're looking for the exact, what is the [00:36:00] rent at this property, that's a fact. That's not a similarity search. So. What we're doing is we're doing an analytical query. It's actually writing SQL queries, hitting an API.
It's actually getting facts, and then the way we've orchestrated it and modeled it, there are fields for facts and there's fields for analysis, and the fields are filled with facts. It's like when you're on any other website airplane, it's not giving you a similarity search for how much this airplane is likely to cost.
It's literally going to a database and pulling out the actual cost for that airplane. So it's a combination of LLMs and also data infrastructure. So that's why it's not gonna hallucinate. It may get confused in the analysis, may say that doesn't seem realistic, but when it pulls from the database, it's giving you a database fact.
Cardiff: It's not gonna introduce a new inaccuracy, even if maybe the analysis is something you agree or disagree with. The analysis or the opinion, it gleans from those facts, but the facts themselves are gonna be reliable, is what you're saying.
Ben: So I'm gonna show you a [00:37:00] couple more things and then we can talk about some of the broader things of what this could mean. So I was gonna do an institutional acquisition analysis of this big apartment building. This 416 units, this built in 1982. Renovated in 2016, it's 98.5% occupied, and I'm gonna say it.
I'd like to buy and upgrade this property. It's called a value add analysis. I want to spend $10,000 a unit replacing countertops, cabinets, and fixtures. This is the type of thing, if you've had an analyst on your team, you say, okay, here's this property. Somebody wants to sell it to us or maybe wanna buy it.
Go do an acquisition analysis and to do that, analyst on your team's gonna want to know some things about it. Like they're gonna say, well, what do you wanna do? Do you would want buy it or do you want to spend some money as even you can upgrade it? So in this case, I told it I wanted to buy it and spend $10,000 on new kitchens and bathroom upgrades.
It understood exactly what that meant, and it came back and asked me, well, what kind of return do you want to achieve? A multiple or your IRR, your ROI? And then also [00:38:00] wants to know my financing assumptions. Am I gonna buy it with any debt? So I'm gonna say I want a 15% annual return and I'm gonna borrow 70% loan to value.
So. These are the types of things you would need, inputs, you would need to do an acquisition analysis. I could have told it a lot more. There's a really good prompt that you probably know card of. You could write a page of directions or the kinds of things you wanna do. So now it's gonna go off and look at current construction innovation cost data from multifamily apartments, gathering recent market data on cap rates, interest rates, loan to value ratios, looking at rent premiums, tenant satisfaction, projected income growth, and valid return assumptions.
So it's doing a lot of thinking as it's going off and doing this analysis.
Cardiff: Is it also looking into how much $10,000 a unit can get you in kitchen cabinets and things of that nature?
Ben: Based on what you normally would spend, if I'd said 25,000, it would've come back with something different. Okay. So it's saying, here's an outline of what I'm gonna go do, and I'm gonna say, does this approach align with my needs? I'm gonna say, yeah, [00:39:00] do this. Yes, this outline looks great. Please do the full acquisition analysis.
Cardiff: It's almost like asking for your permission. It's like we're making these assumptions. Does this work?
Ben: It wants to have some outline of the types of things that, there's so many assumptions you can make and there's a version where you say, no, I wanna tear the building down and build a new building. I mean, there's a lot of different ways you could upgrade this property, so it's now gonna go off. This is a grinding amount of work.
Cardiff: It is a lot. Yeah. Oh, there it is. Legend Oaks value add acquisition analysis, and now it's giving you a deal snapshot, an estimated value, and an investment thesis. All right. Do you wanna take us through the conclusions here?
Ben: There's a lot here. It's gonna produce a lot of information, as you might imagine, like an analyst would, but it is got all the property facts. Here's the outline of essentially what it's saying. It's worth $64 million or $153,000 a unit. To get a 15% IRR at 70% LTVI need to buy it for no more than 62 million.
It's telling me what the risks are. It's doing a property overview. It's [00:40:00] looking at the property metric versus the zip code and the variance. Looking at the operational performance management effectiveness. Tenant profile, market analysis, supply demand, value add strategy. Look, here's where it's saying new countertops, cabinets, refacing, replacing fixtures, bathroom.
Here's how much you're spending. Here's the cost. Benchmarking. A typical kitchen for a full remodel might cost this much. Here's what a bathroom would cost. We're only spending this much eight to 12,000 versus 20 to 40,000. Here's the different phases. Pre-phase, phase one, phase two, and did a full financial analysis looking at the gross rent.
Operating expenses. The NOI, it did a full financial statement, so it's got the rents, the other income, your management fees, your maintenance, and then your taxes and utility. So it gave you your net operating income and it ran a discounted cashflow analysis.
Cardiff: This is great. This means you have some sense of what you, if you're an investor, would bid on something like this to hit those targets. In this case, I think it's telling you to bid 62 million, and if they say [00:41:00] no, then you're not gonna hit those targets and then you're out.
Ben: If you're a financial analyst, one of the most important things you're doing day in and day out, discounted cash flows with Excel spreadsheets. This just produced a discounted cash flow based on the assumptions that I gave it in 60 seconds by year. And if you're a real estate person, that was really easy.
Just a couple more things, then we'll get outta the weeds here and go high level. But often you buy something, you look at the sales comps too, not just the financial underwriting. So you wanna say, okay, how does this look compared to other comparable properties? And then it did a risk assessment matrix of all the major risks like construction costs, lease up delays, interest rate, expansion, the probability of that, the impact and the mitigation strategy pursue, and then the residual risk.
Then looked at market risk execution, risk financing. It gave you a pretty robust real estate analysis in call 90, 120 seconds.
Cardiff: It crunched all the numbers and in 60 seconds gave you something that if you were to print it out and keep it in the same format, it would be like 10 to 15 pages, I'm guessing, of [00:42:00] rigorous analysis here.
Ben: Here's the export to PDF. And another fun thing is see this link, I could share it with you. If I send you that link, you would get this thing, it would look exactly the same, and you could say, Ben is all wet. I want a 20 IRR. So you could start to work on the same analysis I did just by sharing this link so you can share the analysis.
The last thing I'll show you is something really fun as somebody shared with me. We've tried to do, and I think we've done a pretty good job, is we've built the first real estate AI analyst in the world. And I think it can do the job for anybody who's gonna do any kind of analysis. That could be an institutional investor buying and party building.
It could be a small investor, you know, fix and flip, or maybe a single family rental. Or it could be a realtor or a home buyer who's thinking about. I wanna buy this home, help me make a good decision, and it's got data and compute and analysis that is just far more than any one person could possibly do in the time and cost in which we did it.
Cardiff: That's impressive. I would also, to go back to what I was saying earlier about how [00:43:00] researchers could possibly use. Us as well. You could track the trends over time. You could monitor how some of these neighborhoods have changed through the years. You can see the effect of certain macroeconomic shocks and events and how it affects things there too.
Like I think it's a very powerful tool of analysis. Again, I wanna say that for those of you who are listening and not watching it on YouTube, first of all, definitely consider going to check it out on YouTube or just go start playing with the thing itself so you can get the full suite of it. But. I wanna emphasize it again, it's very intuitive to use.
One of the things that's obvious from this demonstration is that it's pretty simple. It's every bit as simple as the homepage of chat, GPT as a starting point at least, and to ask it questions. I think most of the questions you made were mostly in plain human language. Not too many questions that had a lot of real estate jargon until you wanted to, until you wanted to start asking it.
About 15% IRR. And loan to value ratio and things like that, but it's easy for a normal person to start using it. It's not gonna be something that I think is [00:44:00] limited to just being used by investment analysts and specialists in the field.
Ben: I think that today, I mean, we haven't even launched it yet, but today it's not gonna be as good as the best real estate financial. Analyst, I think I would put it at somebody who's two to four years of real estate experience and somebody who's maybe had a year of software programming experience. So not to like toot my own horn, I don't think it's as good as me at making a real estate analytical decision.
So if you're a real estate person who's been doing real estate for 10 or 20 years, it's an analyst that supports you, it's not gonna be better than you. But it's gonna be as good as a first year or third year analyst, and it's gonna be a hell of a lot faster, hell of a lot more comprehensive and a hell of a lot cheaper.
Software's not meant to be better than people who can spend a week thinking about something, but it's a lot faster and a lot cheaper, and that's why it's better. 'cause you could look at a hundred properties in the same time an analyst would take to do one.
Cardiff: And that goes back to actually a longstanding debate about not just ai, but about technology itself, which is that when new and impressive technologies emerge and [00:45:00] everybody's wondering about whether it's going to either replace people or just augment some people who know how to use the technology, and the answer can always be all of the above.
It is a situation where probably you will need fewer of the lower level analysts, the number crunchers, the ones doing the routine parts of the job. It might also really enhance the performance of really well-trained, experienced people who know how to get the most out of a tool like this. And if you've been, in your case in real estate for I think two plus decades, not to date you too much again.
Maybe that chip has sailed, but if you sort of know what you're doing and you know how to combine your own expertise with the data crunching that this can do, then it'll enhance your own performance, even as it eliminates the need for other kinds of tasks. That is the history of productivity growth.
That's how it works when new technologies emerge.
Ben: We're invested in a lot of AI companies, and we use it at Fundrise for all sorts of stuff, but we hadn't built an AI application until now. And when you're actually a [00:46:00] builder, not just a consumer of it, you really learn a lot of its capabilities and its limitations, and there's a. Lot of people have a lot of opinions about ai.
People think it's gonna become Terminator and people think it's gonna be trash. I'm using it. We built something pretty powerful with it. It's amazing to me that it can do as good a job as somebody, really, a person who spent 20 years in high school and college and two or three years in a real estate organization, and then maybe a year learning how to write Python and basic data modeling.
And it can do that job. That's incredible. That's insane. It's doing the job of a person with years of experience. And I think that's gonna be true for real estate. Like all technology, it broadens who can be a real estate investor. It evens the playing field a lot. 'cause this type of technology, I doubt any real estate organization in the country has the infrastructure and sophistication this thing has.
And you can be competitive with like a Starwood or a Blackstone for tens of dollars. That's awesome.
Cardiff: It's one of the things you've talked about a lot, which is democratizing access to financial [00:47:00] markets. This seems to fit into that general approach to the way you do things at Fundrise. Obviously in the. Early Fundrise years and up till now it was in the kinds of funds that you offer to investors. Now it's in a technology that investors can use themselves to better learn this market.
It is a small D Democratic tool.
Ben: Yeah. And then what implications for other AI and AI for lawyers or AI for accounting? You need somebody who knows a lot about the thing like accounting or pick your other kinds of white collar work. But if you can do this real estate, you should be able to do virtually any white collar analysis, white collar knowledge work.
And if it's as good as somebody with a few years of experience and AI is getting better every year, the computes is being dumped into it, data centers, all that stuff. I think that people who are one to five years, even one to 10 years out, it's gonna affect people's ability to get jobs. Then the ones who are more experienced, more curious, more agency think they're gonna be even more [00:48:00] successful.
So the asymmetry of tech is that the best, get most of the benefits. And with ai, I'm seeing it here with real estate, it can make the best real estate investors even better. But you're not gonna hire as many real estate people.
Cardiff: Anything else about the model itself, about the technology that you'd want to emphasize to our listeners before we go?
Ben: It took years to do this. There's so much complexity in it. It's something from a tech point of view. They call it a vertical. So it's for real estate and for it to be good at real estate. We had to do a lot of things that are specific to it. The amount of data we have is incredible. We have billions and billions of data points, billions.
And we're updating it every day. We have data ingestion and pipelines, and we have all this data transformation modeling. I was talking to a guy last week, I was in Los Angeles, and this guy was a quantitative trader. He used to be a quant at one of the big investment banks, and he used to trade and he said, you know, I've seen this movie before.
'cause back in 2009 and 10, Citadel and all these [00:49:00] high frequency traders started trading against me. And I used to have about 250 milliseconds before I could make a trade. Oh, wheat's going down. That means soybeans will go down and the high frequency traders traded in a picosecond. So that opportunity went away and then they started having so much data and so much machine learning, ai, whatever that by 2013, 14, it was like playing poker and they knew what my cards were.
They were so sophisticated. And he's like, I know what this is like. And I was talking to some journalist and she said to me, is this all about the data? I said, no. If you look at Renaissance Tech or Jane Street or Citadel, you wouldn't say, oh, they have more data than me. They have more modeling. They've modeled this data.
There's so much modeling. So that the data has implications and signal. The modeling is something that people who haven't been data scientists, data engineers, wouldn't appreciate how freaking much that matters. And then there's lots of AI work on top AI restoration and engineering, and you're doing this real time.
Those questions were being answered real time anyways. It's wild. So [00:50:00] fun. Yeah, I think that we'll roll it out. I'm hoping, and I can imagine that lots of people will think it's really fun and cool and useful and I think it's gonna make us much better real estate investors already. We use it in our real estate investing and it's gonna make us much better tech investors because we know a lot about the tech.
So it's just such a win-win for Fundrise.
Cardiff: Is one other thing that you wanted to show me related to housing, I think in San Francisco and current events. So why don't you set that up for us?
Ben: So one of our teammates is based in San Francisco. He's an the venture, he's our analyst at our venture fund. He is fabulous. And he asked a question, he said, oh, I wonder how all this money, all these AI companies could be trillion dollars of IPOs and tender offers. How's that gonna affect the housing market in San Francisco?
So he asked real AI that question and it modeled out how the AI boom could affect the housing market and it did a really good job. I wanna go through this fast and maybe I'll put this on the website and anybody can go see this analysis. I'll put it on my Twitter 'cause you can share these links. [00:51:00] So.
Here's a baseline of housing prices. So housing prices pretty flat other than this bubble during COVID pretty flat in San Francisco. So it built a baseline set of assumptions, a baseline set of historical data. It looked at how the Facebook IPO affected home values, which apparently they were up more than 20% in certain neighborhoods. I looked at California IPOs over the last 30 plus years and then it modeled out this one and a half trillion dollars scenario. I'm just gonna get there so much here. Conservative scenario, aggressive scenario. And then he sent it to me and I asked it to graph the model. Those different scenarios out where it shows housing prices going from $1.38 million, a house maybe to two and a half, almost doubling.
And it talks about why, 'cause its supply constraints. There's a number of permits in the entire city, 1000 so knows how many permits in the supply of housing in any market. And then I asked to tell me, well, which neighbors should I buy in? And then went through every neighborhood and recommended Noe Valley, impact Heights, Pacific Heights, which is totally right. [00:52:00] And they told it how much housing prices could be up 50 to a hundred percent in those neighborhoods. Then talked about tier two. Mission Bay in the Castro, and then tier three, Soma and Richmond Sunset. And then it talks about why key drivers buy place. It just went on and on and on, and so much analysis.
And then I asked it to recommend specific homes, which homes should I buy? And it went and it recommended these homes.
Cardiff: Oh, look at the prices on some of these. By the way, for those, for those who are just listening, the first home that it showed up. Up $28 million.
Ben: In Sausalito. So power this home. You sell for 50 million in three years to some AI employee. So it's really good actually doing scenario planning, which is near and dear to my heart. Cardiff, as you know, it doesn't just have backwards looking data. It's really good at doing math and forecasts, and I just thought this was so cool and so timely.
We had no idea that it could do this type of analysis. Because once you have the data and the AI and the modeling and all the tooling, it's ability to [00:53:00] do the work of a person. You ask a person, Hey, real estate analyst, go do this. Interesting thing. I have another one. I'm gonna see if I show to Tyler Cowan around how tariffs are impacting certain cities.
'cause we have a lot of data, so it can do lots of fun stuff that we probably never imagined possible when we originally built it.
Cardiff: It's fascinating, the idea that you could look at places that are specifically affected by certain shocks. In this case, we're talking about a potential AI boom and its effects on certain neighborhoods. But for example, if there's a place that is disproportionately affected by tariffs, then you could see how that might affect home values there.
Housing values.
Ben: You have household incomes, you have fico. Are the people getting delinquent on their credit cards?
Cardiff: I like that it has the multiple path analysis there. You've referred to this before as like scenario planning and it has in fact three different possibilities there. You know, in one case not much changes. They go up a little bit and then a middle case and then the extreme case where housing prices, even in a place in a part of the country where housing's already [00:54:00] outrageously expensive, doubles or even triples.
That's fascinating. And then it does it at the individual level of neighborhoods. And blocks. That's fantastic.
Ben: It's wild to me.
Cardiff: I gotta start saving a lot more money if that Sausalito house is gonna become a real possibility here. Or hope the AI bus comes soon.
Ben: Or you need to invest in some ai.
Cardiff: Either way. Fascinating. And it is available right now. People can go to the website, they can sign up for it, or if not, give us a sense of the timeline of how it's gonna roll out.
Ben: Well, by the time we release this podcast, it will be available, not available at this day. If you went to it, I'd have to whitelist your email address, but I won't release this podcast until we
Cardiff: Until it's ready to go. Okay.
Ben: but we're gonna make it so that you can go on there and try it for free. You can just sign up with an email address and just lemme see if this thing's any good.
'cause most people actually, when I show it to 'em at first, they're like, oh ai, it's all hype this bowl. And then they're like, whoa, okay, wait a second. So we tried to make it really easy for people to test drive it.
Cardiff: Yeah, that's great. [00:55:00] And I think a lot of people really are gonna enjoy it. I know. I can't wait to go check it out once it's widely available. And to close things up, I also just wanna encourage one last time, everybody who's listening along, if you really want to get the most out of this chat that Ben and I just had.
Go check out the video version of this conversation, the YouTube version of it, and then you can tag along as we explored all those different neighborhoods in Tampa, Florida, primarily my hometown and in the different apartment buildings and the different investment possibilities and things like that. I think you really will get the most out of it.
If you check out the video version, but leaving that aside, I hope we've also done a good job for the podcast folks. This was an experiment, I think a successful one, but we'll find out once our listeners get back to us and if they say, yeah, that was fun. Or if they say, are you guys nuts? This is an audio medium.
How did you forget that? Doing a visual demonstration on it. We'll find out. Either way, get in touch with us and let us know. Ben, this was fun.
Ben: Yeah, Carta. Fabulous. Thank you.
Cardiff: You, our listener, have been listening to Onward, the [00:56:00] Fundrise podcast, featuring Ben Miller, CEO of Fundrise. I'm Cardiff Garcia of the Economic Innovation Group and host of the New Bazaar podcast. We invite you again to please send your comments and questions to onward@fundrise.com. And if you like what you heard.
Rate and review us on Apple Podcasts and be sure to follow us wherever you listen to podcasts. Finally, for more information on Fundrise sponsored investment products, including relevant legal disclaimers, please check out our show notes. Thanks so much for listening, and we'll see you next episode.
