Pod Insights The Real Estate Espresso Podcast Saving Thousands on Engineering About this episode Today I want to talk about a very practical use of artificial intelligence, from a real world use case. I’m talking about using a well-written AI prompt to eliminate unnecessary engineering cost by finding existing tested assemblies that already solve the problem. This came up recently in the context of a wood framed floor assembly. The plans examiner at the city rejected the proposed assembly as not meeting the code. Now when that happens, you generally have two ways forward. The first is to use a prescribed assembly, something that is already recognized, already tested, already accepted. The second is to have an engineer design the assembly and stamp it. That second path is often perfectly valid, but it can also cost thousands of dollars. Sometimes many thousands. And if the assembly you want already exists in tested form somewhere in the world, then paying to reinvent it is not adding value. It’s just adding cost and time. If a published result exists, an AI tool can often find it in ways that a conventional search tool could not. --------------**Real Estate Espresso Podcast:** Spotify: [The Real Estate Espresso Podcast](https://open.spotify.com/show/3GvtwRmTq4r3es8cbw8jW0?si=c75ea506a6694ef1) iTunes: [The Real Estate Espresso Podcast](https://podcasts.apple.com/ca/podcast/the-real-estate-espresso-podcast/id1340482613) Website: [www.victorjm.com](http://www.victorjm.com) LinkedIn: [Victor Menasce](http://www.linkedin.com/in/vmenasce) YouTube: [The Real Estate Espresso Podcast](http://www.youtube.com/@victorjmenasce6734) Facebook: [www.facebook.com/realestateespresso](http://www.facebook.com/realestateespresso) Email: [podcast@victorjm.com](mailto:podcast@victorjm.com) **Y Street Capital:** Website: [www.ystreetcapital.com](http://www.ystreetcapital.com) Facebook: [www.facebook.com/YStreetCapital](https://www.facebook.com/YStreetCapital) Instagram: [@ystreetcapital](http://www.instagram.com/ystreetcapital)
Transcript 0:01 Speaker 00 Welcome to the Real Estate Espresso Podcast. Your morning shot of what's new in the world of real estate investing. I'm your host Victor Manash. Today we want to talk about a very practical use of artificial intelligence, one that has nothing to do with hype, but everything that has to do with execution of day-to-day work. I'm talking about using a well-written AI prompt to eliminate unnecessary engineering cost. And this is a real-life case study from our project. Now, I know we're talking a lot about AI on the podcast. This is not an AI podcast. This is absolutely real estate focused. This came up recently in the context of a wood-framed floor assembly. The plans examiner at the city rejected the proposed assembly is not meeting the code, specifically in terms of acoustic performance. Now, when that happens, you have two ways forward. The first is to use
0:46 Speaker 00 a prescribed assembly, something that's already recognized, tested, and accepted within the building code. The second is to have an engineer design the assembly and stamp it. Second path is perfectly valid, but it can also cost thousands of dollars, and sometimes many thousands. And if the assembly that you want already exists in tested form somewhere in the world, you're essentially paying to reinvent it without adding any value. It's just adding cost and time. This is where AI, used properly, can become a force multiplier. Now, it's not that the AI is doing the engineering. It's not. It's that AI is acting as the missing link to connect real information with the folks who need it. And in this case, AI is not making up an answer. That would be dangerous. So what we're doing here is we're scavenging the world for evidence. We're looking for
1:30 Speaker 00 lab test results for manufacturing listings, for published assemblies, for fire reports, all of that testing data that, by the way, an engineer would need to rely upon in order to stamp the drawing. This is not the sort of search that a Google search would turn up a meaningful result. It's a completely different use case. So in this particular case, I created a very simply worded prompt where I described in words, in actual text, the makeup of the assembly, starting with stranded board at the very top, and then engineered joists underneath it, filling the cavity with fiberglass insulation, and then two layers of type X drywall underneath it with resilient channel. Very standard wood-framed floor assembly that's been used in Millions of properties all over the world.
2:12 Speaker 00 So I was right in thinking that I was not breaking any new ground here and was not going to be doing anything that hasn't been done many times before, even though the plans examiner was not accepting it as a standard assembly that would meet the building code. What the AI searched in particular was an assembly from Boise Cascade called Assembly FR2. That assembly matched the proposed structure closely enough to be very relevant, but more importantly, had already been tested to both ASDM and CAN ULCS 101 for fire resistance. Those are recognized standards. So in other words, the issue might not have been that the assembly was invalid. The issue was that the assembly hadn't been put in front of the examiner in a form that they could accept. Really turns out to be a documentation problem, not a design problem.
2:55 Speaker 00 A lot of money gets spent in development, not because the problem is technically difficult, but because the information's disorganized, it's incomplete, or it's inaccessible. But most importantly, the plans examiner wants to know who's liable for the decision and whose liability insurance is going to cover it if something goes wrong, because definitely is not going to be the plans examiner's liability insurance. Now, I want to be clear, this is not a license to bypass engineering when engineering is required. In fact, we have an acoustics engineer within our team, and I sat down with our acoustics engineer and said, okay, Patrick, let's engineer this assembly, let's get the engineering stamp on it and get that in front of the plans examiner. And his response back to me was, yes, we can do that, but before we do that, we need to find test results for that specific assembly so that we can put a stamp in front of it.
3:40 Speaker 00 Otherwise, we will need to do the testing. And it was that very sentence that was the clue for me that if the information exists, maybe the test results are sufficient. So instead of asking who can design this for me, the question was, has this already been designed, tested, and accepted somewhere else? It's a much more valuable question. A weak question would be, what's the acoustic performance of this floor assembly? But instead, I said, find the lab-tested results for this assembly that show the acoustic performance. By asking the question that way, I'm asking for evidence, I'm not asking for opinion. I'm not asking for the AI tool to synthesize something out of nothing. I'm asking it simply to find the test results. I'm looking for sources, not summaries.
4:20 Speaker 00 So the epilogue to this story is we sent the test results with all of the files from Boise Cascade to the plans examiner and he accepted the assembly with no further questions. That completely eliminated the need to do any engineering work whatsoever. I don't know what the exact savings would be, but it was certainly in the thousands of dollars. You don't hire an engineering team for nothing. And the fact is, the world of construction these days is extremely technical. It's not like it was back in the day when people just walked around with a hammer and put houses together. In order to get assemblies accepted by the plans examiner and by the building inspector, they have to meet fairly high standards. In some ways, this is the use of AI as a slightly more sophisticated search tool, which is not really harnessing the complete power of AI at all.
5:02 Speaker 00 But it's connecting the dots in a very meaningful way that previous search technologies couldn't do. It's only possible by starting to search for the right things. For the building owner, for the constructor, for the operator, judgment still matters. In fact, it matters even more. Someone needs to know what to ask. They need to evaluate whether the assembly is truly comparable. You still gotta put the evidence in a coherent package and someone still needs to decide whether the match is good enough or whether it crosses the line into requiring fresh design work. AI doesn't remove that responsibility, but it does remove search friction and that's a big deal. As you think about that, ask yourself this question, in your own business, how many professional fees are being spent solving a knowledge problem rather than a real technical problem? That's where the disciplined use of AI could be of tremendous value.
5:47 Speaker 00 As you think about that, have an awesome rest of your day. Go make some great things happen. We'll talk to you again tomorrow.
Previous Hospitality Held Hostage Next How Robo Taxis Will Change Real Estate Design Page updated November 14, 2026 · transcript and summary are machine-generated