Latest / Net Effective / Treating AI like a junior: building trust in AI

Transcript

Treating AI like a junior: building trust in AI · Net Effective

• 0:00
  SPEAKER_01: According to Claude, I'm the smartest guy alive.
• 0:02
  SPEAKER_01: Every question I ask, I'm a genius.
• 0:04
  SPEAKER_01: You have to kind of set yourself up to where it's like, no, I want you to be hard on me.
• 0:08
  SPEAKER_01: It tells me all my ideas are good, and I know they're not.
• 0:11
  SPEAKER_01: I'm not trying to replace people.
• 0:13
  SPEAKER_01: I'm trying to replace tasks that really can be done by a machine much faster and hopefully more accurately.
• 0:20
  SPEAKER_00: That was Adam Siegel, and I'm Jonas Bordeaux, host of NetEffective.
• 0:25
  SPEAKER_00: Welcome back for another week.
• 0:26
  SPEAKER_00: Adam, thrilled to have you on the show.
• 0:27
  SPEAKER_00: Why don't you kick us off by uh telling us a little bit about yourself and the organization you work for?
• 0:32
  SPEAKER_01: Absolutely.
• 0:32
  SPEAKER_01: Well, thank you for having me, Jonas.
• 0:34
  SPEAKER_01: You know, we talk all the time at shows when we see each other, and uh I really appreciate being a client of Dwellzy.
• 0:40
  SPEAKER_01: My name is Adam Siegel.
• 0:41
  SPEAKER_01: I'm the vice president of product growth at Crexy.
• 0:43
  SPEAKER_01: A lot of my job is creating relationships with companies like Dwellzy and others to bring the best data into our platform and really find solutions so that we can be the all-encompassing sort of tip of the spear for anything commercial real estate for people when they're either trying to sell a property, buy a property, or do research on their market or their industry.
• 1:04
  SPEAKER_01: We want to be the first thing that people think of.
• 1:06
  SPEAKER_01: So yeah, that's kind of what we do.
• 1:08
  SPEAKER_01: And we live and breathe uh commercial real estate every day.
• 1:11
  SPEAKER_01: And we've certainly gotten caught up, as some have, uh, with the AI bug.
• 1:15
  SPEAKER_01: And uh we're starting to build a lot of tools to help people using AI as people are starting to navigate those waters.
• 1:21
  SPEAKER_00: So thanks for having me.
• 1:23
  SPEAKER_00: Well, it's absolutely a pleasure, Adam.
• 1:24
  SPEAKER_00: One of the things I was really excited for this conversation for is that you and your colleagues over at Craxy probably talk to more transactor in the market every week than just about anyone else I know.
• 1:34
  SPEAKER_00: And I'm so curious to hear how are those folks using AI?
• 1:39
  SPEAKER_00: This is such a relationship business.
• 1:42
  SPEAKER_00: It's so much about, you know, knowing the right people, getting the information that's not published to be able to transact.
• 1:48
  SPEAKER_00: But how are they in that world leveraging AI?
• 1:52
  SPEAKER_00: I think everybody's excited about AI.
• 1:54
  SPEAKER_01: And I think most people would agree that technology is one of the things that is often overlooked in commercial real estate.
• 2:00
  SPEAKER_01: It's more of an expense than an investment.
• 2:02
  SPEAKER_01: People kind of cringe a little bit when they have to write the check because they're not sure they're getting the full value out of it.
• 2:08
  SPEAKER_01: But what I've seen with AI and what I think my colleagues have seen is AI is very exciting to everybody, whether you're talking to a broker or you're talking to an appraiser or a lender.
• 2:18
  SPEAKER_01: And certainly for the private investor, it's opening up a lot of opportunities and it's somewhat leveling the playing field.
• 2:25
  SPEAKER_01: At least that's what people seem to feel, is that they now have access to things more quickly, more data, and they're able to do things at a velocity that they weren't able to do before, whether that's looking at a lot of deals and seeing if it fits their buy box, they're able to understand markets better, whether that's market rents, what a market is doing, what new construction.
• 2:46
  SPEAKER_01: They're all able to do that with a few prompts and really get accurate answers to their questions without having to spend hours and hours digging around and going through their email to look at reports and dig into a bunch of different sites and go down the rabbit hole.
• 3:01
  SPEAKER_01: And I think that is the one thing that AI is really good about is it's gonna create speed within the industry.
• 3:07
  SPEAKER_01: And there's a saying that we often use, which is time kills deals.
• 3:11
  SPEAKER_01: And the more we can get, more velocity we can pick up is gonna just increase the overall value of the market we feel by getting these deals done quickly and breaking down those walls around data, as you kind of touched upon.
• 3:24
  SPEAKER_00: I'm hearing a couple of things from that.
• 3:25
  SPEAKER_00: A lot of them around speed, not only speed to confidence around a transaction you might want to do, but also speed of activity so you can spend more time on the highest value items.
• 3:38
  SPEAKER_00: Is that accurate?
• 3:39
  SPEAKER_00: Am I hearing you right?
• 3:40
  SPEAKER_00: You're nailing it, Jonas.
• 3:41
  SPEAKER_01: A lot of this industry is a lot of repetitive tasks that can be automated, whether it's through workflows or just agents going out and doing things on a regular cadence.
• 3:52
  SPEAKER_01: And I think that helps improve the overall relationship time that you have to be able to create these relationships, create these bonds, do the little bit extra to make sure the deal comes your way.
• 4:04
  SPEAKER_01: And I think what's exciting about all of this is that really, for years, you saw a lot of siloed information.
• 4:12
  SPEAKER_01: You know, everything was housed in somebody's inbox or in their desktop.
• 4:16
  SPEAKER_01: And now people are able to go out and do a much deeper level of thinking very easily.
• 4:21
  SPEAKER_01: They're not having to hire a team, an army of analysts or executive assistants or researchers.
• 4:26
  SPEAKER_01: They're able to do this on their own.
• 4:28
  SPEAKER_01: And I think that's very powerful for people to be able to do this work in a very timely manner.
• 4:33
  SPEAKER_01: I think there's still a confidence issue and a trust issue.
• 4:36
  SPEAKER_01: You touched upon that a little bit.
• 4:37
  SPEAKER_01: And I think that will get better over time, but we're still in the very early innings of AI within the industry in the prop tech side.
• 4:44
  SPEAKER_01: And I think if you give it another year or two years, you're gonna see a ton of gains, a ton of efficiency, and you're just gonna start to see that speed come to market and people able to transact a lot more quickly.
• 4:56
  SPEAKER_00: That's great to hear.
• 4:56
  SPEAKER_00: Certainly, everybody needs more time.
• 4:59
  SPEAKER_00: Uh, that is, you know, the only commodity, as many people say.
• 5:02
  SPEAKER_00: So I'm curious, Adam, how are people actually getting that out there?
• 5:06
  SPEAKER_00: You know, obviously there are purpose-built tools like Cracksea that are now much more AI enabled than they've ever been before.
• 5:12
  SPEAKER_00: And so, you know, AI tools that give that speed are being embedded in different prop tech solutions.
• 5:18
  SPEAKER_00: But then there's the general purpose tools like Cloud or Chat GPT.
• 5:23
  SPEAKER_00: And then there's the kind of tools that cross all industries, whether it's you know, CRM or contact tracking systems.
• 5:29
  SPEAKER_00: What are you actually seeing those transactor using to get the value?
• 5:34
  SPEAKER_01: We're starting to see people really want to understand how they can capture the data more efficiently and then extract that data in a way that fits what they're trying to do, but also be able to have that ability to recall that information at a later date, right?
• 5:52
  SPEAKER_01: Because that information is still valuable today.
• 5:55
  SPEAKER_01: It's valuable a year from now as well, and it's valuable probably even a couple of years from now.
• 6:00
  SPEAKER_01: Certainly there's an expiration date on some of these as we we talk about rental growth or we talk about what lease rates are.
• 6:08
  SPEAKER_01: That's always ever changing.
• 6:10
  SPEAKER_01: But it's still good to have these points in time and see what velocity or what transaction types happened over time.
• 6:17
  SPEAKER_01: And I think being able to organize large data sets for people who maybe are not analytics people, that's gonna be a huge game changer for people to have that.
• 6:27
  SPEAKER_01: The other thing is that breaking down those siloed walls of how data is housed is really something that's gonna bring that speed and efficiency to the industry.
• 6:36
  SPEAKER_01: I think when we look at AI, we look at how can we build workflows that will really help our customers, our clients, be able to do what they need to do quickly, efficiently, and repetitively.
• 6:48
  SPEAKER_01: They're not having to rebuild it every time.
• 6:50
  SPEAKER_01: I think a lot of people are trying to build with AI, whether that's at a company level or even at a personal level.
• 6:56
  SPEAKER_01: And I'm interested to see because there is a lot of time, I'm sure you've built Jonas with our French friend Claude or some GPT or whatnot.
• 7:04
  SPEAKER_01: Yes, in theory, you can start to get there and you can maybe even get 80 or 90% of the way there.
• 7:11
  SPEAKER_01: But it's not always perfect, it's not always elegant, and it's usually built very much for you.
• 7:16
  SPEAKER_01: You can't hand it off to somebody else and they can pick it up very easily to use it.
• 7:21
  SPEAKER_01: And I think that's where there's going to be sort of a little bit of a gap from a user experience, meaning you can use it, but I couldn't use it.
• 7:29
  SPEAKER_01: I wouldn't understand your methodology.
• 7:32
  SPEAKER_01: And I think that's where software companies like us can kind of create with a little bit broader brush strokes things that work for everybody and then allow them to sort of drill down that extra layer.
• 7:44
  SPEAKER_01: So you start here, that's gonna be for 80, 90% of people, but the people who do want to take it to the next level, we can do that.
• 7:50
  SPEAKER_01: And that could be whether it's exposing our data through an MCP server or allowing you to tie directly into Cloud and use Crexy data within that and tie that in with other data sets.
• 8:00
  SPEAKER_01: I think it's gonna be interesting to see how that evolves.
• 8:03
  SPEAKER_00: Yeah.
• 8:04
  SPEAKER_00: Yeah, it is really interesting to think about that all of that data, because there's a tremendous amount of data that a transaction professional is going to deal with as they underwrite deals, deal specific information that's unique to that specific potential transaction, but then also, you know, times a hundred, times a thousand, however many deals they're thinking about a year.
• 8:22
  SPEAKER_00: I think about my friends in SFR who might underwrite 20,000, 30,000 properties a month.
• 8:27
  SPEAKER_00: The amount of information that they have to absorb is enormous.
• 8:30
  SPEAKER_00: And being able to master the tools that allow them to do that is fantastic.
• 8:34
  SPEAKER_00: And kudos to Crexy for making an MCP available and other techniques that allow people to interact with it.
• 8:40
  SPEAKER_00: So I have a question about that for you.
• 8:42
  SPEAKER_00: One of my friends in the industry called MCP a scam last week.
• 8:46
  SPEAKER_00: Does that mean anything to you?
• 8:47
  SPEAKER_00: I was I was kind of mystified by that.
• 8:49
  SPEAKER_00: Dwellzy has an MCP.
• 8:50
  SPEAKER_00: We have clients starting to use it to interact with our data.
• 8:53
  SPEAKER_00: Any idea why he would have said that?
• 8:54
  SPEAKER_01: I guess I'd have to understand why he feels that.
• 8:57
  SPEAKER_01: What event triggered that thought?
• 9:00
  SPEAKER_01: Because that seems like something that there was a triggering event, meaning he had expectations perhaps of what the MCP was going to deliver.
• 9:06
  SPEAKER_01: Maybe it didn't meet that because it wasn't exposing the level of data.
• 9:09
  SPEAKER_01: And I think that's what everybody's trying to figure out, right?
• 9:12
  SPEAKER_01: Is that you don't want to give away everything, but you want to make it a little bit easier for people to use the tools they're comfortable with.
• 9:20
  SPEAKER_01: I think the industry as a whole is morphing a little bit to where not everybody wants to work on your platform.
• 9:27
  SPEAKER_01: They want to be able to take that data easily, other places, and manipulate it.
• 9:30
  SPEAKER_01: And you're not going to be able to give them all the tools.
• 9:33
  SPEAKER_01: But certainly I'd want to ask that question like, what was it that the MCP didn't do for you?
• 9:38
  SPEAKER_01: Because I think that is one of those things that you start at level A or level one.
• 9:43
  SPEAKER_01: And level one is not an MCP.
• 9:45
  SPEAKER_01: That's more like level three or four in your AI journey.
• 9:48
  SPEAKER_01: So if he was expecting something like it was this magical crystal ball that could just solve all his problems, maybe it didn't meet his expectations.
• 9:56
  SPEAKER_01: But I think for people who really want to manipulate the data, have access to the data and be able to keep drilling down, not be satisfied with just the question and answer, but really kind of push back and forth and have that sort of tension back and forth between the data and the problem you're trying to solve.
• 10:11
  SPEAKER_01: I think that's where an MCP server really helps you.
• 10:14
  SPEAKER_01: Because I know myself back in my early days as an analyst, I would build canned reports.
• 10:19
  SPEAKER_01: And then I realized the canned report just leads to more questions.
• 10:22
  SPEAKER_01: And I would have to go back.
• 10:23
  SPEAKER_01: And then I realized that no, I'm much better with raw data in Excel and a pivot table.
• 10:27
  SPEAKER_01: I think it's kind of the same thing, right?
• 10:29
  SPEAKER_01: For basic things, you're going to want this.
• 10:31
  SPEAKER_01: And then for other things, you're going to want to be able to interact with the data in a way that maybe requires a little bit different level and skill.
• 10:38
  SPEAKER_00: Yeah, I think where he was coming from is as a vendor of data, as a vendor of software, he was worried that the LLMs were going to be able to capture his data, perhaps.
• 10:47
  SPEAKER_00: But it was an interesting perspective that I hadn't heard before and really got me thinking about our own use of MCP.
• 10:53
  SPEAKER_00: And I think trust in all of this is a huge challenge for us to think about.
• 10:58
  SPEAKER_01: Yeah, but I don't think that makes it a scam, right?
• 11:01
  SPEAKER_01: I mean, I think it's more of a worrisome, hey, how do I protect my data moat, right?
• 11:05
  SPEAKER_01: Which we always talk about here.
• 11:06
  SPEAKER_01: I know you and I have talked about data moats in the past.
• 11:09
  SPEAKER_01: And Crexy, we have a large marketplace that gives us some of our data that is our moat, just like you have your data moat.
• 11:15
  SPEAKER_01: Again, if you're building off of that and you have a data moat, yes, you don't want to just give away the farm, but there is some value in allowing people to have easier access to what they need.
• 11:25
  SPEAKER_01: Because when you create that friction, I think the friction is what causes people to go elsewhere.
• 11:30
  SPEAKER_01: I want people to use Crexy in a way that's very, very impactful for them on a day-to-day basis.
• 11:35
  SPEAKER_01: I want Crexi to become a daily habit for people where they start their day and end their day and anywhere in between that they need us, we're there to serve them, whether it's with answers, whether it's with them finding their next deal on our marketplace.
• 11:47
  SPEAKER_01: I want to be able to service everybody when they need me.
• 11:51
  SPEAKER_01: And hopefully that builds over time, right?
• 11:53
  SPEAKER_01: I want people using us all the time.
• 11:55
  SPEAKER_01: I think that to say it's a scam would be a little bit rough, but certainly I also worry.
• 11:59
  SPEAKER_01: I worry today, right?
• 12:01
  SPEAKER_01: There's bots, I'm sure, crawling your site, there's bots that crawl my site.
• 12:04
  SPEAKER_01: We're dealing with agents that are creating accounts and doing all this nefarious sort of things.
• 12:09
  SPEAKER_01: It's a new world we live in, new age we live in for sure.
• 12:12
  SPEAKER_00: Yeah.
• 12:12
  SPEAKER_00: No, if you told me the amount of time I put into preventing thieves from stealing data, it's absurd.
• 12:18
  SPEAKER_00: Not something I wanted to be spending my time doing.
• 12:19
  SPEAKER_00: It's frustrating to you, right?
• 12:21
  SPEAKER_01: Because it's always the bad actors first, right?
• 12:23
  SPEAKER_01: And it causes a ripple-down effect.
• 12:25
  SPEAKER_01: And we deal with it, Jonas.
• 12:27
  SPEAKER_01: Trust me, I wish I had a magic wand that that would be the one problem I would probably try to solve.
• 12:30
  SPEAKER_00: Yeah.
• 12:31
  SPEAKER_01: Because I want to make it safer out there for everybody.
• 12:33
  SPEAKER_01: But you hit on the word trust, and there's got to be trust in the data.
• 12:37
  SPEAKER_01: There's got to be trust in the answers.
• 12:39
  SPEAKER_01: We've all experienced it where, according to Claude, I'm the smartest guy alive.
• 12:43
  SPEAKER_01: Like every question I ask, I'm a genius.
• 12:45
  SPEAKER_01: And you know, that sort of give and take when it's like everything, you're just getting this affirmation that what you're doing is the right thing because it's burning tokens.
• 12:54
  SPEAKER_01: You have to kind of set yourself up to where it's like, no, I want you to be hard on me.
• 12:58
  SPEAKER_01: Is this something I should do?
• 12:59
  SPEAKER_01: Is this something I shouldn't do?
• 13:00
  SPEAKER_01: Is why has this never been done before?
• 13:03
  SPEAKER_01: I think people are still figuring that out.
• 13:05
  SPEAKER_01: And I think that's where it gets a little scary to me because it tells me all my ideas are good and I know they're not.
• 13:10
  SPEAKER_00: Yeah.
• 13:11
  SPEAKER_00: Have you ever asked Claude to tell you about yourself?
• 13:14
  SPEAKER_00: Uh somebody told me to do that.
• 13:15
  SPEAKER_01: I think it knows enough about the crazy stuff I ask it already that I'm afraid that it it's it knows way too much about me.
• 13:22
  SPEAKER_00: I haven't met anyone who hasn't been told that they are a notable local leader so far.
• 13:28
  SPEAKER_00: Everybody, every one of us is a notable local leader.
• 13:32
  SPEAKER_01: You know, I laugh because I see people who get really, oh, I want, I don't want the internet to know anything about me.
• 13:38
  SPEAKER_01: And I don't, I don't want my iPhone to listen.
• 13:40
  SPEAKER_01: It's like, okay, then you need to go live somewhere in Wyoming or North Dakota and you need to live off the crib.
• 13:46
  SPEAKER_01: Because unfortunately, we that ship sailed a long time ago.
• 13:49
  SPEAKER_01: If you have a driver's license, if you have an iPhone, if you have anything like that, then the world knows about you.
• 13:56
  SPEAKER_01: And I think that's the same thing that unfortunately hasn't quite reached our industry yet to understand that that data is available.
• 14:05
  SPEAKER_01: You might have to dig a little bit harder for it.
• 14:07
  SPEAKER_01: But what people need to understand is one, the data has a shelf life, but also by keeping it so close to the best, it doesn't guarantee you're gonna be able to monetize that.
• 14:17
  SPEAKER_01: And if you really want to make money, you need to be understand that being more open is probably gonna benefit you better than being closed off.
• 14:27
  SPEAKER_01: And you know, I could give historical examples of that throughout history of people who wanted to close off their borders and how that worked out for them as a society.
• 14:34
  SPEAKER_01: But it's kind of the same thing with data right now.
• 14:37
  SPEAKER_01: You know, if we all had access to data, we could make better decisions, which means I may buy your property or I may understand what I could do here at this particular location.
• 14:47
  SPEAKER_01: It's not being that sort of closed-off entity is not gonna survive very well much longer, in my opinion.
• 14:54
  SPEAKER_01: I could be completely wrong.
• 14:55
  SPEAKER_00: So, you know, a lot of transaction professionals that I've talked to in the past have always thought that their unique access to data, their knowledge of things that nobody else knows is vital to their success.
• 15:09
  SPEAKER_00: Do you think that has to change?
• 15:10
  SPEAKER_00: Do you think that's changing?
• 15:12
  SPEAKER_00: I do.
• 15:13
  SPEAKER_01: I think it has to change simply because if that's the case, it has to be that yes, I have this piece of knowledge.
• 15:20
  SPEAKER_01: But unless I go out and broadcast that I have that piece of knowledge, which then everybody knows because I've broadcast it, I have to wait.
• 15:27
  SPEAKER_01: It's almost like just sitting in the dark and waiting for the light to come on.
• 15:31
  SPEAKER_01: If somebody doesn't flip the light switch on, you can't actually have.
• 15:35
  SPEAKER_01: That's the interesting thing, is that people feel like, I know what this comp is, I know what this sold for, I know what this, unless someone's asking me, there's really no value in that.
• 15:43
  SPEAKER_01: And then because somebody asks me, how do I make that valuable to them to where it actually puts money in my pocket or gets me a listing?
• 15:52
  SPEAKER_01: It's almost like a needle in a haystack, right?
• 15:54
  SPEAKER_01: You have this piece of data and somebody has to be looking for that particular needle that you're holding on to.
• 15:59
  SPEAKER_00: It's like the mythical off-market perfect deal that you know everybody's excited about finding or doing.
• 16:05
  SPEAKER_00: And it's kind of impossible to find that and keep it a secret.
• 16:09
  SPEAKER_00: So you only you have it.
• 16:10
  SPEAKER_01: Right.
• 16:10
  SPEAKER_01: I think you're hitting it exactly right.
• 16:12
  SPEAKER_01: Is that by being able to expose that, you have more people who are likely to come to you and ask for your advice.
• 16:18
  SPEAKER_01: You're gonna show that you're knowledgeable, you're gonna show that you're an expert in the field, and that probably benefits you more in the long run.
• 16:24
  SPEAKER_01: It's gonna be a fundamental change for a lot of people who've been in this industry for any amount of time.
• 16:29
  SPEAKER_01: So if you've been in the industry for 10 years, you were trained that you just held your data close to the vest.
• 16:34
  SPEAKER_01: But I think a lot of the younger people who are coming into this are gonna see what Claude and GPT and AI can do for them.
• 16:40
  SPEAKER_01: They may not have that sort of, I need to hold everything as close to the best.
• 16:44
  SPEAKER_01: I actually am better off sharing and being more open because it's gonna actually win me more business in the long run.
• 16:50
  SPEAKER_00: Yeah.
• 16:51
  SPEAKER_00: I do see a baseline assumption from the next generation after ours that that people are gonna know, that there's always somebody in the know that what you're doing is not private, that fundamentally, you know, they're used to sharing on social media and they're used to seeing other people share, you know, more than our generation might have been comfortable with on social media.
• 17:10
  SPEAKER_00: And so that reflects then in their professional lives.
• 17:13
  SPEAKER_01: For sure.
• 17:13
  SPEAKER_01: And again, I think that in today's day and age, people almost expect everything at their fingertips.
• 17:19
  SPEAKER_01: So the expectation is that data is available.
• 17:22
  SPEAKER_01: And I think that's going to become more and more the norm, not the exception.
• 17:26
  SPEAKER_01: I just don't know if that's two years from now, five years from now, or 10 years from now, but it's definitely going to happen.
• 17:33
  SPEAKER_01: And I think, like I said, it goes back to the whole, it's going to benefit everybody with greater velocity as far as how quickly transactions could get done, which is going to benefit everybody in the long run, right?
• 17:44
  SPEAKER_01: Because again, I'm not taking as long to get a deal under contract.
• 17:47
  SPEAKER_01: I'm not taking as long to get it through due diligence and get it closed.
• 17:50
  SPEAKER_01: So I can move on to the next deal.
• 17:52
  SPEAKER_01: And again, it's we live in a world of instant gratification, right?
• 17:55
  SPEAKER_01: Everybody has Amazon and they want it.
• 17:57
  SPEAKER_01: Like there's no reason why deals should take as long as they do.
• 18:01
  SPEAKER_01: And if we can cut 10%, 20%, if it goes from 60 days down to 45 days, because we're able to get information parsed through more quickly, then we've done our job.
• 18:12
  SPEAKER_01: And that's what we want to do at Corexy.
• 18:14
  SPEAKER_01: And I'm sure that's what you want to do at DwellZ is have people be able to get through what they're trying to get through quickly so that they can move on to the next thing.
• 18:24
  SPEAKER_00: Yeah, we talk a lot about uh speed to confidence.
• 18:26
  SPEAKER_00: Yeah.
• 18:26
  SPEAKER_00: You know, how can we help people accelerate the path to a confident decision?
• 18:31
  SPEAKER_00: There's an interesting overlay, which is I was looking at a piece of software a couple of weeks ago that a partner's putting together.
• 18:37
  SPEAKER_00: And they were able to gather most of the information that in my experience would have taken six weeks to three months to assemble.
• 18:46
  SPEAKER_00: And they put together that package in less than 15 minutes for somebody.
• 18:50
  SPEAKER_00: But I don't think anybody can absorb it and get to confidence on a decision in that kind of time, even if it is perhaps better quality information, but delivered at that speed.
• 19:00
  SPEAKER_00: How can the human brain get to confidence about a multimillion dollar decision in 15 minutes when it's used to taking six to 12 weeks?
• 19:08
  SPEAKER_01: And that's the same thing that we've always found with things like valuation and underwriting.
• 19:12
  SPEAKER_01: Even if I give you the full due diligence package, I give you every lease, every amendment.
• 19:17
  SPEAKER_01: And again, some asset classes are easier, right?
• 19:20
  SPEAKER_01: Maybe multifamily is easier because they're all one-year leases and you really just need a rent roll and understand, and you need to understand what the market is from a are rents moving up, are rents moving down, what can I accurately do uh or forecast?
• 19:33
  SPEAKER_01: But in other asset classes, certainly, you know, there are leases that are have multiple amendments.
• 19:37
  SPEAKER_01: Do you feel confident that the AI has extracted all that information, not missed a clause, not missed a cap on cam expenses, or that there's leakage here, or that there are clauses there where you know you can't have tenants, certain tenants in the thing, and you're hoping to lease a vacant space.
• 19:55
  SPEAKER_01: I don't think I would have confidence today, but I know that the velocity, because again, even two years ago, I look at from January of 2025 to now, call that 18 months.
• 20:07
  SPEAKER_01: The models have gotten better, the tools have gotten better, it's gotten smarter.
• 20:11
  SPEAKER_01: Yeah, I agree, Jonas, there is a confidence level that we need to get to that this is doing the job as well as my analyst was doing it before.
• 20:20
  SPEAKER_01: But I think we're getting pretty close to that pretty quickly.
• 20:23
  SPEAKER_00: Yeah.
• 20:23
  SPEAKER_01: I mean, I think we're still maybe a year or two away on the underwriting side.
• 20:27
  SPEAKER_01: But there's gonna be other environmental.
• 20:29
  SPEAKER_01: I don't want to miss an environmental thing, you know.
• 20:31
  SPEAKER_01: I don't want to miss something that's gonna cost me a lot because I know I have to dig up the site or something like that.
• 20:36
  SPEAKER_00: Or of course, yeah, all that stuff.
• 20:38
  SPEAKER_00: No, definitely.
• 20:39
  SPEAKER_00: You have to have a complete understanding and an accurate understanding.
• 20:43
  SPEAKER_00: And if we're getting the AI feeding us incorrect information, that's gonna be problematic.
• 20:47
  SPEAKER_01: You don't want to be the first, and this will happen.
• 20:49
  SPEAKER_01: I eventually we will see an article out there where somebody bought an asset.
• 20:54
  SPEAKER_01: It might be for$2 million, it might be$20 million, and they used AI, and AI missed something, and they made a big mistake, and now they're gonna be sitting having to go back to investors or go back to their boss, their investment committee, and say, we made a$5 million,$10 million mistake because the AI missed this.
• 21:10
  SPEAKER_01: And yeah, that person will probably lose their job, but it it'll be a very hard lesson to learn.
• 21:15
  SPEAKER_01: But nobody wants to be that first person to make that mistake, at least publicly.
• 21:21
  SPEAKER_01: I'm sure people have made mistakes already, but I don't think anybody wants to make that mistake.
• 21:25
  SPEAKER_01: And like I said, there's a very real possibility that happens.
• 21:29
  SPEAKER_00: Yeah, which is interesting.
• 21:30
  SPEAKER_00: When that happens, there's gonna be a lot of hand-wringing and there's gonna be a lot of like, oh, thank God, that's not me.
• 21:34
  SPEAKER_00: But the reality is I can list off the top of my head any number of transactions that have gone sideways because somebody missed something.
• 21:41
  SPEAKER_00: Correct.
• 21:42
  SPEAKER_00: And fundamentally, this is just a different flavor of that.
• 21:45
  SPEAKER_00: But again, back to getting to confidence, when the process of due diligence has fundamentally changed, how do you review that due diligence in an intelligent way?
• 21:56
  SPEAKER_00: Because you're not absorbing it in the same way at the same pace with that.
• 22:00
  SPEAKER_00: That same level of iterative interaction that you may have in the past.
• 22:04
  SPEAKER_00: So, how do you underwrite all of that due diligence effectively?
• 22:08
  SPEAKER_00: How do you review it all effectively?
• 22:10
  SPEAKER_01: So for me, I think that you have to take it for what it is.
• 22:13
  SPEAKER_01: So I assume that if I'm using AI, it is as if I just hired somebody new.
• 22:17
  SPEAKER_01: They don't really know what they don't know yet, and they are junior, and I have to watch and review everything they do.
• 22:23
  SPEAKER_01: And as their confidence level grows and they get better skills because they've seen more and more things, I get more confident that maybe I don't have to review as much.
• 22:32
  SPEAKER_01: But it's that confidence level, to be honest, that really is the biggest gap right now.
• 22:38
  SPEAKER_01: Because I mean, honestly, a junior analyst, you're right, a junior analyst makes mistakes.
• 22:43
  SPEAKER_01: But am I able to find that mistake more quickly because the AI can do it in 10 minutes or 15 minutes versus a junior analyst that's taking a day and then coming back to me?
• 22:53
  SPEAKER_01: And that gets back to the whole speed and time kills deals sort of analogy.
• 22:58
  SPEAKER_01: Would I rather know that it be able to catch a mistake and know that it's gonna work 24-7 for me?
• 23:03
  SPEAKER_01: And it's gonna be able to, if I if I leave at five o'clock and I tell it I want to review on my desk by seven the next morning, the agent is going to be able to do that.
• 23:10
  SPEAKER_01: Even if it has some mistakes, I'm gonna have that.
• 23:12
  SPEAKER_01: I'm not gonna have that two and three days later.
• 23:15
  SPEAKER_01: That's where it gets exciting for me is that this is where I'm not trying to replace people.
• 23:19
  SPEAKER_01: I'm trying to replace tasks that really can be done by a machine much faster and hopefully more accurately over time as it gets more ups.
• 23:29
  SPEAKER_01: What I get frustrated with is that we see all the time, I think you probably see it all the time as well, is we see these sort of new startups that are solving one issue and they tell you that they're solving it a hundred percent accurately.
• 23:41
  SPEAKER_01: And you and I both know that that's just not necessarily possible to do, i.e., lease perfect lease abstraction.
• 23:49
  SPEAKER_01: It's like, no, that's just because what happens when you get the handwritten lease?
• 23:53
  SPEAKER_01: I've seen those, like where people are still writing handwritten leases, yeah, or you're missing something, right?
• 23:58
  SPEAKER_01: You're missing an amendment and it just assumes.
• 24:00
  SPEAKER_01: And we've all seen the hallucinations with AI, right?
• 24:03
  SPEAKER_01: It's like, oh, well, we just assume that number because we couldn't find it.
• 24:06
  SPEAKER_01: It's like you can't just assume a number.
• 24:08
  SPEAKER_01: A number's a number because you need the actual number.
• 24:10
  SPEAKER_01: So that's what I get a little frustrated with is that some of these people come out with we have the perfect solution and you know, it's ready in 10 minutes.
• 24:19
  SPEAKER_01: And it's like, okay, maybe it is.
• 24:21
  SPEAKER_01: And again, I'm not saying that people can't do that, but it just I worry.
• 24:25
  SPEAKER_01: I worry that people are gonna sign up for things and expect things, and that's where where everybody's gonna get a bad taste in their mouth with AI.
• 24:31
  SPEAKER_01: Because that also affects me and you, right?
• 24:33
  SPEAKER_01: Because people who've had bad experience before, I'm having to actually convince them twice as hard that my product has been vetted and used by tens of thousands of people.
• 24:43
  SPEAKER_01: And, you know, I have millions of people on my website every month.
• 24:45
  SPEAKER_01: Like that's where I have to do even a better job of convincing them.
• 24:49
  SPEAKER_01: And that's that's hard because they've been burnt.
• 24:51
  SPEAKER_00: Yeah.
• 24:51
  SPEAKER_00: Well, that's always the way when we're dealing with a new technology.
• 24:54
  SPEAKER_00: Like I've been watching the development of autonomous driving pretty closely.
• 24:58
  SPEAKER_00: And, you know, I remember when one of the first traffic accidents happened.
• 25:03
  SPEAKER_00: And, you know, there's thousands, tens of thousands of traffic accidents every day driven by humans.
• 25:07
  SPEAKER_00: And the first time a machine caused one, people lost their minds.
• 25:11
  SPEAKER_00: You know, uh, how can we possibly be doing this?
• 25:13
  SPEAKER_00: You know, and you know, so I think it's just it's something new for folks to absorb, and they have to get used to a new source of insecurity, if you will.
• 25:22
  SPEAKER_01: Yeah.
• 25:23
  SPEAKER_01: I mean, from your standpoint, Jonas, what's the biggest thing that is keeping people from adopting AI at an even faster rate in this industry?
• 25:32
  SPEAKER_00: What I'm seeing is people being what I might call just gun shy.
• 25:37
  SPEAKER_00: I don't know where to start.
• 25:39
  SPEAKER_00: And people have downloaded ChatGPT on their phone and they're asking it questions, you know, having it substitute for Google for them, but they don't know what's next after that.
• 25:51
  SPEAKER_00: That I think is the biggest stopping point.
• 25:54
  SPEAKER_00: So I'd actually turn that around to you.
• 25:56
  SPEAKER_00: Like, what do you think is next?
• 25:57
  SPEAKER_00: If you're a transaction professional, what's the stack that you could use to get started with AI that would really be net effective, that would be accretive to that transaction professional?
• 26:09
  SPEAKER_01: I think it's some of the stuff that we're starting to release that we have in beta right now, which is we're really looking at workflows that will help people on a daily basis.
• 26:18
  SPEAKER_01: And we're looking at that in a way that says whether you're a single person who works as a one-man or one-woman shop within the industry, or you're part of a small team, or you're part of a you know, a large institution, we want to build workflows that help on a daily basis.
• 26:34
  SPEAKER_01: So there's a daily need.
• 26:35
  SPEAKER_01: And that can be anything from a research assistant that's using Crexy proprietary data.
• 26:40
  SPEAKER_01: You're welcome to bring your own data in and let us kind of morph it all together for you and be able to have really interactive discussions and really drill down.
• 26:48
  SPEAKER_01: Our latest evolution of our research AI tool can now literally drill down to a street in a neighborhood and tell you what's going on at the street level, where before you were waiting for the Los Angeles multifamily or such and such report to be released by a CBRE or a JLL.
• 27:04
  SPEAKER_01: And those research teams do a phenomenal job.
• 27:07
  SPEAKER_01: But really, people want to get much more granular today because you've been to Los Angeles.
• 27:11
  SPEAKER_01: There's a big difference between the South Bay and Venice and Santa Monica, and then go literally a mile in and you're into Inglewood, completely different markets, but they're all Los Angeles.
• 27:22
  SPEAKER_01: And as you start to look at that, you really want to be able to know this is where I want to be, this is where I want to buy, this is where I already own, I want to understand.
• 27:29
  SPEAKER_01: Because going a mile away, the rents could change dramatically.
• 27:32
  SPEAKER_01: And again, it's that type of granularity.
• 27:34
  SPEAKER_01: It's almost like that's expected now, is that people want that instant gratification to be able to do that.
• 27:39
  SPEAKER_01: And the data's out there.
• 27:40
  SPEAKER_01: You just have to be able to do the deep research.
• 27:43
  SPEAKER_01: And I think the biggest thing for us is tuning, right?
• 27:46
  SPEAKER_01: Tuning our agents to really be commercial real estate focused.
• 27:50
  SPEAKER_01: Yes, anybody can go to Claude, but Claude's not going to understand it as well as we do because they don't have that data mode that we have.
• 27:56
  SPEAKER_01: I think that's where our sort of focus is going to be is how do we create things like, you know, we have a product called Create, which helps you create BOVs and offering memorandums, and eventually you'll be able to create websites and email marketing so that you're doing things once, but getting four different results from that one set of data that you're uploading.
• 28:17
  SPEAKER_01: So you can start by building your BOV, jump in your due diligence, build your BOV, you win the deal, it goes straight to an OM.
• 28:23
  SPEAKER_01: The OM is approved, you click a button, it creates a website, it posts a correct sea.
• 28:28
  SPEAKER_01: Again, saving time, even if it's an hour here, 15 minutes here, three hours here.
• 28:32
  SPEAKER_01: It all adds up.
• 28:34
  SPEAKER_01: But also it's starting to know you, like you said.
• 28:36
  SPEAKER_01: It's starting to really understand what you do and really be able to drive value by giving you insights that maybe you didn't see.
• 28:43
  SPEAKER_01: I think that's where the other opportunity for AI is, is that AI is all knowing and it's gonna see things that you may not see or be aware of things that you may not be aware of.
• 28:52
  SPEAKER_01: And it can bring it to your attention in a meaningful way.
• 28:55
  SPEAKER_00: Yeah.
• 28:55
  SPEAKER_00: No, that's one of my favorite uses is giving it a large amount of information and saying, what's wrong with this, or what are the exceptions, or what are the outliers in all of this?
• 29:04
  SPEAKER_00: Uh it's just it's so much better at finding those than I am.
• 29:06
  SPEAKER_00: Oh uh, the ability to sort through a thousand pages of documents about a property, I can't do that effectively.
• 29:12
  SPEAKER_00: Uh, but it can.
• 29:13
  SPEAKER_00: No, and again, you miss things, right?
• 29:15
  SPEAKER_01: You fat finger things.
• 29:16
  SPEAKER_01: It can do that too.
• 29:17
  SPEAKER_01: Don't get me wrong.
• 29:18
  SPEAKER_01: But it's a lot better to sit back and try and pick holes at what the AI came back with because I've found that when I push on the AI and say, are you really sure here?
• 29:26
  SPEAKER_01: It'll go back and think.
• 29:27
  SPEAKER_01: Again, I go back to it's a junior analyst.
• 29:30
  SPEAKER_01: It's gonna need the ups, the reps, whatever you want to call it, to get better.
• 29:35
  SPEAKER_01: But I have seen it learn over time, learn about me, learn about what I'm doing, be able to reference projects that I did two months ago and say, this sounds a lot like what we were building before.
• 29:45
  SPEAKER_01: Do you want me to pull that up?
• 29:46
  SPEAKER_01: Be like, no, but I want you to use these pieces.
• 29:48
  SPEAKER_01: And it's like, it's a partner, like it's an assistant that starts to know me.
• 29:52
  SPEAKER_01: And I think that's where it gets really interesting, right?
• 29:55
  SPEAKER_01: Because imagine if you had that trusted partner for five years and you had five years worth of things within your Claude account and it can pull back and have that elastic recall.
• 30:05
  SPEAKER_01: Man, that's that's some pretty abundant intelligence that's sitting at your fingertips.
• 30:08
  SPEAKER_01: If you just now get in the daily habit of, I'm gonna dump everything.
• 30:12
  SPEAKER_01: I'm gonna let it see my inbox.
• 30:14
  SPEAKER_01: I'm gonna let it understand what I'm doing.
• 30:16
  SPEAKER_01: But there's that trust factor.
• 30:17
  SPEAKER_01: I think it all comes down to trust.
• 30:19
  SPEAKER_01: If you have the trust, you'll figure it out.
• 30:21
  SPEAKER_00: You'll figure it out.
• 30:22
  SPEAKER_00: Yeah.
• 30:22
  SPEAKER_00: So I hear you on using great tools that folks are building within systems like Craxy.
• 30:27
  SPEAKER_00: For somebody who's nervous about getting started, other than using those tools and you know, the extended capabilities of systems that they're already familiar with and using, what would you suggest to get them started to get them comfortable, you know, thinking about the transaction professional?
• 30:41
  SPEAKER_01: You know, I think you have to come with a problem that you're trying to solve, right?
• 30:46
  SPEAKER_01: So you have a need.
• 30:47
  SPEAKER_01: There's some pain point in your life that could be solved with AI and give AI a chance.
• 30:53
  SPEAKER_01: You will probably be surprised at the result.
• 30:56
  SPEAKER_01: Good or bad, but probably more good than bad.
• 30:59
  SPEAKER_01: And if it can solve that problem, then you start to, it's almost like getting you get bit, right?
• 31:04
  SPEAKER_01: And you get excited.
• 31:05
  SPEAKER_01: You want to start to use it.
• 31:06
  SPEAKER_01: And you're like, oh, it did that really well.
• 31:08
  SPEAKER_01: I wonder if it could do this.
• 31:09
  SPEAKER_01: I wonder if I could try this.
• 31:11
  SPEAKER_01: And by the way, it doesn't mean that it all has to be done within, meaning you don't have to build it all yourself.
• 31:16
  SPEAKER_01: But if you have confidence that, wow, this could really make a difference in my life, you can start to look for those tools that are built on AI, right?
• 31:25
  SPEAKER_01: Or have that component to it.
• 31:27
  SPEAKER_01: I think it's still gonna be a hill to climb, Jonas.
• 31:29
  SPEAKER_01: I think, like I said, we're in the very early innings of what AI can mean to this industry.
• 31:34
  SPEAKER_01: And I think there's gonna be certain asset types that will adopt a lot more quickly.
• 31:38
  SPEAKER_01: I think if I look at multifamily, it's primed for AI to help in a lot of different ways, especially a small, small operator who maybe owns 20, 30, 40, 50 doors.
• 31:49
  SPEAKER_01: It can help a lot, right?
• 31:51
  SPEAKER_01: Not just in acquisition or disposition, but understanding what rents are.
• 31:54
  SPEAKER_01: But even with things like tracking maintenance that I need to do, and there's a lot more that can be done than just acquiring deals.
• 32:01
  SPEAKER_01: There's property management aspects, there's collections, there's all the sort of things that can be done and use AI as a tool within that industry.
• 32:09
  SPEAKER_01: As you go into other asset classes, it may be a little bit slower to adopt, but I think that there's no industry I can think of, or no asset class I can think of that wouldn't benefit from AI in some way, shape, or form.
• 32:21
  SPEAKER_00: I completely hear you.
• 32:22
  SPEAKER_00: Changing gears a little bit.
• 32:24
  SPEAKER_00: How are you using AI in your personal life?
• 32:26
  SPEAKER_01: So for me, it's interesting.
• 32:28
  SPEAKER_01: I think AI for me is when I have an idea for Crexy, whether it's a new feature or a new tool or whatnot, I can very easily mock it up and get a really interesting, working, clickable model.
• 32:40
  SPEAKER_01: I can even tie that into real Crexie live data, right?
• 32:42
  SPEAKER_01: So I can tie into Snowflake and I can pull live data and I can make this look and feel.
• 32:47
  SPEAKER_01: Again, I can then hand it off like we talked about earlier.
• 32:50
  SPEAKER_01: I can then hand it off to real developers because I'm not a real developer.
• 32:53
  SPEAKER_01: I can hand it off to them and say, look, this is how I view it.
• 32:56
  SPEAKER_01: Or I can hand it off to my UX designer and say, hey, how do you envision this working within the Crexy system today?
• 33:03
  SPEAKER_01: And they have no problem pushing back, but at least I'm giving them a head start, right?
• 33:06
  SPEAKER_01: I'm not doing mock-ups and PowerPoint.
• 33:08
  SPEAKER_01: I'm not doing even Figma, right?
• 33:10
  SPEAKER_01: That's kind of a thing of the past.
• 33:12
  SPEAKER_01: But I think that's the cool thing, right?
• 33:14
  SPEAKER_01: Is that I can work it to a point where I've shaped it enough and now I can hand it off to somebody, not just, hey, Jonas, I'm thinking about this.
• 33:21
  SPEAKER_01: It's a button that over here.
• 33:22
  SPEAKER_01: And when you click the button, like, and and we're having to have two, three, four, five hour conversations to get to a clickable prototype and then the back and forth, right?
• 33:30
  SPEAKER_01: I can hand them something that's my vision and then they can take it and run with it.
• 33:34
  SPEAKER_01: And I can literally hand them the project, right?
• 33:36
  SPEAKER_01: All the files, everything else.
• 33:38
  SPEAKER_01: For me personally, it's interesting.
• 33:40
  SPEAKER_01: I think there's a lot of ways that people can make money using AI if they're if they're thinking ahead and they try.
• 33:46
  SPEAKER_01: And I think it's worth trying, right?
• 33:48
  SPEAKER_01: Because you might hit on something that could change your life, could make you money, whether it's a second revenue stream or something like that.
• 33:54
  SPEAKER_01: I think it's very interesting because again, it can parse so much data and it can see trends and it can see things, and it can help you evaluate a business and if you wanted to buy a business, and it can help you build all these sort of tools, like a business plan and PLs and how to track all this thing.
• 34:11
  SPEAKER_01: I think it opens up a lot of possibilities for people.
• 34:14
  SPEAKER_01: But for me on my daily basis, I'm I'm using it to sort of experiment with things that could be within Crexie, using our data, seeing what other type of data is out there, seeing how we could tie that all together.
• 34:24
  SPEAKER_01: And it gets it gets really interesting.
• 34:26
  SPEAKER_01: It's a lot of fun.
• 34:26
  SPEAKER_01: I burned through a lot of credits at tokens, but that's luckily my token bill is paid for by the company.
• 34:32
  SPEAKER_01: But yeah, it's uh it's fun.
• 34:35
  SPEAKER_00: It's yeah, that's fantastic.
• 34:37
  SPEAKER_00: Yeah, you can get yourself in trouble with token use.
• 34:39
  SPEAKER_01: It's one of the fringe benefits of being an employee at Crexy, and and they encourage it.
• 34:42
  SPEAKER_01: Like that's the whole thing.
• 34:43
  SPEAKER_01: You just never know.
• 34:44
  SPEAKER_01: And and we strongly have encouraged for the last 18 months for everybody to get on the AI train.
• 34:49
  SPEAKER_01: And our engineers got on the train first, and they have adopted it very quickly, and they're using it, and it's really a force multiplier.
• 34:57
  SPEAKER_01: But even our product managers and our designers are now using it more and more and more because those lines get a little blended and they're really starting to build some cool stuff.
• 35:06
  SPEAKER_01: I'm so excited with what our folks are building.
• 35:08
  SPEAKER_01: And like I said, I think it's just power because we can move fast, and that's what you want.
• 35:13
  SPEAKER_01: You want to be able to move fast, iterate, get things out to market, test it, see if people like it.
• 35:18
  SPEAKER_01: You don't want to take six months, put it out there for two months, and then find out it was you want to get that answer much much more quickly.
• 35:24
  SPEAKER_01: So you can go back to the drawing board if you have to and get v2, v3, v4.
• 35:28
  SPEAKER_00: So yeah, 100%.
• 35:30
  SPEAKER_00: Well, Adam, in each episode to try to recognize uh somebody in the industry who's doing a fantastic job using AI.
• 35:37
  SPEAKER_00: Anyone you comes to mind for you?
• 35:39
  SPEAKER_00: You want to recognize?
• 35:40
  SPEAKER_01: From an AI standpoint in this industry, I really am impressed with what our engineers do.
• 35:45
  SPEAKER_01: Our front end, our back end, even our VPs here are all in on AI.
• 35:50
  SPEAKER_01: And they're doing a great job every day of, again, using it as a force multiplier, but also really coming up with some creative solutions, really sort of rebuilding the stack to be AI forward and understanding where the industry is gone.
• 36:02
  SPEAKER_01: And it's kudos to our CPO and our CTO who have that foresight to really use the AI and really was a top-down approach of pushing this and saying, hey, if not, we're gonna be left behind and we're just gonna be a marketplace.
• 36:15
  SPEAKER_01: And we didn't want to just be a marketplace.
• 36:17
  SPEAKER_01: Uh, it's a nice business, but we want to be an AI company uh and we want to bring AI solutions to the industry.
• 36:24
  SPEAKER_01: And like I said, we have two and a half to three million users a month who are visiting our site.
• 36:29
  SPEAKER_01: We certainly can be that on-ramp for AI for a lot of people in the industry and not just help ourselves, but help others who are trying to build similar AI tools.
• 36:38
  SPEAKER_01: We can be that on-ramp for people.
• 36:40
  SPEAKER_00: Well, Adam, this has been such a pleasure.
• 36:42
  SPEAKER_00: Thank you so much for coming on the podcast.
• 36:44
  SPEAKER_00: Really appreciate you.
• 36:46
  SPEAKER_00: This has been Adam Siegel on NetEffective.
• 36:49
  SPEAKER_00: I'm Jonas Bordeaux.
• 36:50
  SPEAKER_00: Please, wherever you're listening to this podcast, please like and subscribe and go to netaffective.show to sign up for our email so you can find out when new episodes are coming out.
• 37:01
  SPEAKER_00: Really appreciate all of you listeners.
• 37:02
  SPEAKER_00: Uh, any questions, you can always find me via LinkedIn.
• 37:05
  SPEAKER_00: Adam, how can folks reach you?
• 37:06
  SPEAKER_00: They can reach me from LinkedIn as well.
• 37:08
  SPEAKER_01: It's uh Adam Siegel.
• 37:09
  SPEAKER_01: You'll find me under the Crexie.
• 37:11
  SPEAKER_01: Happy to connect with anybody, happy to answer questions, and happy to let people know what we're building to see if it would help them.
• 37:16
  SPEAKER_00: All right.
• 37:16
  SPEAKER_01: Thanks again, Adam.
• 37:17
  SPEAKER_01: Thanks, Jonas.
• 37:18
  SPEAKER_01: Appreciate it.