AI Is Coming for SaaS in Proptech

As baseball Hall of Famer Satchel Paige once wisely advised, “Don’t look back — something might be gaining on you.”

Such might be the recommendation to proptech software as a service (SaaS) providers who are dealing with strong competition from artificial intelligence companies in the field. The game, though, ain’t over until it’s over, as another Hall of Famer, Yogi Berra, sagely noted.

While AI is challenging SaaS for technological dominance in real estate, most signs point to at least a period of peaceful coexistence before the possibility that the former supplants the latter.

“I don’t think AI is killing SaaS entirely,” said Zach Aarons, co-founder and general partner at MetaProp, a Manhattan-based early stage proptech venture capital firm. “I find it hard to believe that the majority of companies are going to rip out their systems of record and replace them with either their own vibe-coded system of record, or that they’re going to bypass the need to have a system of record at all.”

Vibe coding allows companies to build technology products without knowing themselves how to code by using existing technology such as Anthropic’s Claude, Aarons explained. “Without prior programming education and knowledge you can build tremendously complex web tools,” he said. “We’re seeing this across our portfolio. Two years ago across our portfolio companies, adoption of these tools was at 30 percent. Now it’s at 100 percent.”

Right now, companies that want to build a basic customer relations management (CMR) system without any underlying technology to replicate a real estate deal-sourcing CRM do not need a technologist behind it, said Aarons. All that is needed is someone knowledgeable about product architecture to make it look and feel usable to the rest of the organization.

However, the future of vibe coding is somewhat problematic.

“In seven years, we’re going to be in a real pickle unless the AI coding tools have gotten so good that they’re perfect, because no one’s hiring junior engineers right now,” Aarons said. “No one needs them to write code. So you have existing mid- to high-level engineers who learned how to code in the pre-vibe-coding era who are now knowledgeable enough to babysit the AI. In seven years, unless the AI gets exponentially better and you don’t have to babysit it, there won’t be anybody to babysit it because you’ll have senior engineers retiring out of the workforce. It’s like what’s happening with plumbers. We always say that for every two plumbers leaving the workforce, there’s only one plumber to take their place. That’s kind of what we’re seeing across computer programming.”

David Stifter, founder of PredictAP, a Boston-based AI-powered SaaS platform for invoice ingestion and coding in commercial real estate, sees a different wrinkle in the AI-SaaS relationship.

“I think it’s not really SaaS that AI is going after, but more entry-level jobs,” said Stifter. “For example, call center work or things like that. The thing that’s different about AI is that with traditional tools you couldn’t answer knowledge-based real estate questions like, ‘Hey, what units are available?’ or ’What’s my price for this or that service?’

“What we do is invoice coding in real estate, and there’s a lot of complexity around the accounting of how do I do the accrual for a specific bill? It’s very nuanced customer- or tenant-specific knowledge that traditionally someone would learn over the months and years of doing their job. And it’s those knowledge-based jobs that are being more threatened by AI than say the SaaS platforms, where the core accounting is happening already. There is still going to be an accounting book of record.”

A key question in the AI versus SaaS landscape is whether incumbent real estate firms are afraid of moving away from SaaS because they are too entrenched in that technology and can’t disengage enough to move to AI.

“That is the absolute question,” Stifter said. “I think on one hand, they are afraid. EliseAI is interesting because they started as a chat company, and then they realized, ‘Well, we’re talking to the tenants, so why don’t we become a CRM company?’ Once you start to build knowledge, information and relationships, the ability to touch more and more things is much easier, and you don’t have that legacy system that’s fairly difficult to change.”

Real estate companies deciding between using AI or SaaS, or a blending of both, reside on a wide spectrum, said Vijay Mehra, CEO and founder at LenderBox, an AI-powered commercial real estate lending platform.

“There are some companies that are fully committed to investing in AI and replacing everything that they’re used to using — traditionally SaaS platforms — with AI,” said Mehra. “Another thing that you’re seeing today is the speed to market with technology development. Once upon a time, it would take you six months to a year to go and build a platform, whether it’s a SaaS platform or any sort of solution to solve a problem. Now you can do that in large part in a week.”

That’s true at his own firm, Mehra said. Everything built internally at LenderBox would once have used traditional management tools from a Salesforce.com or a HubSpot, he said. 

“We’re now building these tools in-house in weeks,” Mehra said. “So, now, we don’t have these heavy SaaS subscriptions and any of the costs associated with it. And I think a lot of commercial real estate firms are actually realizing the same thing, especially those that are sophisticated and armed with good technology talent.

“But, then, on the other end of the spectrum, we’re selling into banks, into private credit teams, and credit unions,” he added. “A lot of these folks are just now starting to adopt SaaS, which is kind of a crazy scenario, thinking how far along we actually have come with AI. So you’re still seeing that gamut or spectrum of use. Most of these traditional SaaS platforms don’t have that AI intelligence layer we’re providing, so, depending on the level of how much they want to embrace, it’s either going to complement existing SaaS infrastructure, or it’s going to rip and replace. Either way, companies are realizing the value that is being added is undeniable, and they have to adopt it in some form or fashion.”

Mehra doesn’t see AI killing SaaS. Instead, he believes we will see a blended model, where companies stack technologies on top of their existing infrastructure, trying to have the best of both worlds in the foreseeable future. “I think there’s sort of a knee-jerk reaction right now where the pendulum is swinging too far to one side, but I think there’s going to be some normalization that happens.”

Sandeep Ahuja, CEO at Cove, an AI-driven platform for architects and designers, strongly disagrees.

“We literally are AI killing SaaS in proptech,” Ahuja said. “Developers are very tuned in to what AI can do for them. I was talking to Rukus Esi, the chief digital officer at [multifamily giant] Avalon Bay, and he talked about how they’re already leveraging AI in the back office to make workflows so much better. Now they’re thinking, ‘OK, where else? Where else? Where else?’ And architecture seems to be such a natural spot, that when we go in with the messaging, ‘Hey, we can cut your time to construction drawings by about 50 percent,’ they’re like, ‘Yes, yes. We have seen AI do that in other parts of our lives already.’”

The speed of AI iteration is another compelling factor in its growing use by companies, said Patrick Chopson, co-founder and principal at Cove.

“People don’t realize that AI is doubling in capability every seven months,” said Chopson. “So when people say, ‘Oh, this will never happen.’ Why? I mean, to me, I’m looking at it from the coding side of things as well, and we can see that now the machines are almost better than 99 percent of software engineers. They code better than them. That’s coming for all professions. Our work is going to be fundamentally different.”

Other proptech experts found some nuances in the AI versus SaaS discussion.

“AI isn’t killing SaaS in proptech, it’s making it smarter,” Nihar Malik, chief innovation officer at MRI Software, said in an email. “In commercial real estate, core systems of record, such as property management, lease accounting [and] facilities and financial operations, are as essential as ever. These platforms hold a company’s most critical data, and that data layer is the fuel for AI. You can’t separate intelligence from the systems that power it. The importance of SaaS isn’t changing, but how people interact with it is. AI is becoming an intelligence layer across trusted platforms, enabling conversational access to data, automating reconciliations, flagging anomalies and surfacing insights before a user even asks.” 

The shift, then, according to Malik, isn’t away from SaaS, but from static software to intelligent, connected platforms. 

“So, no, SaaS companies aren’t disappearing because of AI,” he said. “If anything, the opposite is happening.”

Mike Sroka, CEO and co-founder of Dealpath, summed up a lot of the yin and yang with AI and SaaS, agreeing with the primary point that both will continue in differently evolving forms.

“AI isn’t killing SaaS — AI is changing what ‘good SaaS’ looks like,” Sroka said in a statement. “In real estate, customers don’t buy software for software’s sake; they buy outcomes (speed, precision, auditability, collaboration). The winners will be platforms that combine workflow depth with high-quality structured data and AI orchestration, not AI bolted onto thin workflows.

“We’ve seen consolidation and repositioning, but ‘AI replacing SaaS’ is often overstated. More commonly, AI features are being embedded into existing systems, and point solutions face pressure if they can’t integrate into a system of record.”

As for the money chasing after AI and SaaS, Sroka sees capital having clearly rotated toward AI narratives, “but durable value still accrues to platforms with distribution, retention and data moats — especially those that can productize AI into repeatable workflows.”

Philip Russo can be reached at prusso@commercialobserver.com.