According to a recent JLL survey, 92% of CRE teams have started piloting AI, or plan to start this year, with McKinsey estimating that AI could generate $110-180 billion in value for the real estate sector. Yet despite this widespread interest, actual deep adoption remains surprisingly low. The gap between awareness and meaningful implementation is what Topher has spent the last 2.5 years working to close.
Key topics covered:
The same 2025 JLL survey reveals the recipe for successful CRE AI adoption with three main ingredients: mature data infrastructure, established change management processes, and experienced teams. Unstructured and scattered data is the first roadblock and, unfortunately, the most common one.
“ From what we see here at Ascendix, the highest ROI is very consistent in document automation, like proposals. We even do our proposals, reports, packages, and summaries with the help of AI tools. But data and data structuring are crucial.
AI adoption looks different at the company vs individual levels. While things like document processing, proposal generation, and information extraction from documents are, quote, “super powerful for almost any company,” they might not always be the best AI for commercial real estate use cases for all professionals.
“ You create an AI automation or a custom GPT that's automatically going to put together their LOIs for them and save them time? Yeah, you can totally do that. That's not all that difficult to put together. Do you want to give that to a junior broker who hasn't typed up 50 LOIs yet? Probably not, because when are they actually going to learn to put together an LOI? ”
“ “Some brokers come up to me, and they're like, 'Hey, I work at CBRE, or I work at this giant firm, we got a fleet of people making brochures. I don't really need that technology.' But then I'll have five other people come up to me and be like, 'Hey, I work at a three-person shop. We're stuck doing our own brochures all the time. This is going to change my life. I spend hours and hours per week just putting together my own brochures. Now I can do that in a couple of minutes using AI, and it looks better than what I was putting together before.'” ”
Deep research is an AI feature available in major platforms like ChatGPT, Claude, and others that can synthesize information from multiple sources across the web. For commercial real estate, Topher identified several key applications like market research, deal sourcing, tenant prospecting, finding distressed properties, etc.
“ Deep research just has so many implications in commercial real estate. Everyone should at least experiment with Deep Research and look into some ways that it can save you time. ”
“ A huge topic in commercial real estate is just decks, you know, people spend so much time on proposals, on marketing materials, on brochures. And even if you don't need AI to work on your brochures, just think of all the other marketing materials that you're putting out there into the world, you know, that just don't look that good. ”
Recommended AI tools for marketing materials generation:
Notebook LM is Google’s AI research tool that grounds responses only in the sources you provide it. This solves a critical problem: hallucinations and made-up information.
“ Not everybody has a giant interconnected database with an AI agent hooked up to it that can go in and retrieve all the files that it needs at the right time.
Notebook LM is a really convenient solution for that. It's AI that is grounded only in the sources that you provide it. So if you want some information from a lease agreement, or all the lease agreements for your building, or your rent roll, because you want to figure out what tenants are coming up for renewal, if that information is inside Notebook LM, you can just query that information and get whatever insights you want. And you can do it very quickly.
When you want to synthesize that information in different ways, you can pop it into Notebook LM and create a podcast for yourself that you can listen to on the way to your next site visit or the next tour. ”
Notebook LM is not the only solution for centralised data processing. Arthur has been working with AI assistants and automation tools for various organisations whose systems have different infrastructures.
“ We were considering Notebook LM that is a part of the Google Workspace ecosystem. So if you’re working with Gmail and Google Drive, then Notebook LM is a great tool for you. If you’re working on the Microsoft stack, like with Outlook and Teams, then Microsoft Copilot would be a better solution just because it has access to all the data already. ”
If your commercial real estate firm operates on Salesforce, there’s a native solution that delivers similar functionality: Agentforce. Salesforce’s AI agent platform provides the same document analysis, data synthesis, and intelligent automation capabilities you’d get from Notebook LM or Microsoft Copilot—but fully integrated with your existing Salesforce environment.
Agentforce can access your CRM data, communication history, property records, and deal pipelines without requiring data migration or switching between systems. For CRE firms already invested in Salesforce infrastructure, this means AI capabilities work directly within the workflows your team already uses daily, eliminating the friction of adopting yet another tool in an already complex tech stack.
With our domain expertise and Agentforce Quick Start Package, you’ll have have a functional AI agent within 1 month.
“ I think the more impactful technology that is going to kind of change the lives of people in commercial real estate is these automation platforms. Because everyone loves talking about AI agents. You’ve got to build an AI agent. That word is so overutilized that it’s lost all meaning. Most people have no idea what an AI agent actually is anymore. ”
Agentic coding is a software development approach where autonomous AI agents plan, write, test, and debug code with minimal human intervention.
Arthur shares behind the scenes of one of the ongoing Ascendix’s projects that has shown an 80% cost reduction thanks to an agentic coding approach. This is what Arthur calls the “democratization” of technology: tools that were once only accessible to Fortune 500 companies are now available to boutique firms at a fraction of the cost.
“ A year ago, a custom project with high volume data processing and complex logic, might easily be a $5 million engagement. And today, with agentic coding, the same scope can be delivered for less than $1 million. That’s not hype. This is the project that we’re already doing. ”
A December 2025 study titled “Professional Software Developers Don’t Vibe, They Control” found that experienced developers value agents as productivity boosters while insisting on fundamental software quality attributes and employing strategies to control agent behavior.
Organizations that rushed in hoping to replace developers or operate on “set and forget” autonomous systems discovered this approach doesn’t work in practice. Instead, over 90% of engineering leaders plan to expand their use of AI coding tools as assistants integrated into workflows, not as replacements.
“ Do you think there is a danger of vibe coding and people just saying, 'Hey, I’m going to vibe code my own SaaS products, and I’m going to stop paying companies Ascendix?' ”
“ The short answer is yes, of course. The vibe coding is a thing… But you need to have some skills. And these skills are mostly about the ability to analyze the business and the problems, to create instructions for agents that you use for vibe coding. And actually understanding the coding itself on the architectural level. ”
Critical considerations for vibe coding:
Security is the biggest concern of the vibe coding approach. Data encryption, secure document transfer, and protection from malware (malicious software) are not something you can ask an AI agent to create and continuously monitor. With increased usage of AI for cyberattacks, relying on the same AI agents for protection may result in data leakage, pricy ransom payment to restore access to your system, and irreparable reputation damage.
“ We have developed numerous AI applications inside our company to automate, speed up, and increase the productivity of our teams in different departments. But the core, the infrastructure that we use to support our business, is reliable, secure systems. We don't vibe code them. ”
TL;DR
Topher outlined three main reasons for slow technology adoption in CRE:
“ [CRE] is not such a giant industry, but if you try to make something that works for property managers, it's probably going to have very little utility for a broker. Then you say, “Okay, we're going to be something that serves commercial real estate brokers.” There's more nuance there, too. Are you talking about leasing people? Are you talking about investment sales people? Are you talking about tenant reps?
“ When you try to roll something out to a company, there’s going to be some younger folks that just pick it up like lightning. And then there’s going to be some folks that are closer to the top who might have a little bit more trouble with it. And it's not that old people don't want to use technology, but sometimes there can be some hurdles there. ”
“ It's difficult for the SaaS companies to get it right, but they promise that they are getting it right. And then a commercial real estate group spent all this time migrating over, and they realized it was not the right decision. And they're already too fatigued from that transition to want to even try to do it again, to try the thing that might actually be the right solution. ”
Ascendix Technologies has spent over 20 years specializing exclusively in commercial real estate technology to understand broker workflows, property management needs, and capital markets from the ground up. With 300+ clients globally including JLL, CRESA, Transwestern, and Savills, Ascendix combines deep domain expertise with technical mastery.
Also, Ascendix Technologies provides ongoing Concierge support to ensure successful adoption—because the right technology without proper implementation and training still leads to the same migration fatigue Topher described.
Book a Free Consulting Call with our experts and find out how AI can benefit your organisation.
First, let’s define “adoption.” Most real estate professionals have heard about AI and even used popular LLMs like ChatGPT or Claude. But can it really be called adoption? And the main problem, in Topher’s opinion, is that “way fewer people are experimenting with it enough to actually find the most valuable components of AI for their role.”
Despite these challenges, Topher sees a fundamental shift happening:
“ “Everybody’s excited about technology in commercial real estate in a way that I’ve seen for the first time, because AI has made technology so much more accessible to so many more people. ”
AI adoption in commercial real estate has moved from “nice to have” to a competitive necessity. The PropTech market is projected to expand from $34 billion in 2023 to $90 billion by 2032, driven largely by AI capabilities.
The question is no longer whether to adopt AI in commercial real estate, but how quickly you can find the right tools, training, and workflows for your specific needs.
The Concrete Voice is a commercial real estate technology podcast started by Arthur Ambartsumyan after over 10 years of honing his expertise in CRE, PropTech, FinTech, AI, and other technologies. He asks thought-provoking questions, so you get insightful answers and practical tips. The main goal is to make our listeners more informed about CRE technology innovations, trends, and future possibilities.
Listen to The Concrete Voice podcast on YouTube, Apple Podcasts, Spotify.
We choose our guests very carefully to ensure the quality of insights but we are always open to collaboration. Please, contact us via email [email protected] to discuss the opportunity for your appearance in the next episodes.
Kateryna is an experienced writer focused on topics related to Salesforce consulting and optimizing operations within the real estate industry via CRM implementation. Through her insightful articles, Kateryna aims to help organizations make the most of their Salesforce adoption.
(26 votes, average: 5.00 out of 5)
Loading...