The commercial real estate investment market is opening up. Multifamily sales volume reached $22.8 billion in the first quarter of 2026, cap rates have stabilized in the mid-5% range, and refinancing pressure is surfacing selective entry points for well-capitalized buyers rather than broad distress. For investors, the setup is favorable, and it rewards those who can evaluate the most opportunities the fastest.
That last part is the catch. More deals worth underwriting means more offering memorandums, rent rolls, and operating statements landing on analysts’ desks. Before anyone can decide whether a 200-unit apartment complex or a suburban office repositioning is worth pursuing, someone has to pull the financial data out of the documents and get it into a model. The firms that compress that step get to the judgment calls sooner, and in a market where capital is disciplined and selective, getting to the judgment call first is often the difference between winning the deal and reviewing it a day too late.
David Bratslavsky, the founder of QuickData.ai, built his company around that bottleneck. His argument is not that AI should replace the analyst’s judgment. It is that AI should eliminate the hours of mechanical work that prevent analysts from using their judgment sooner.
“Pricing has stabilized, so the edge isn’t guessing where cap rates go anymore,” Bratslavsky says. “The edge is coverage. Every hour an analyst spends on data entry is a deal your competitor underwrote, and you didn’t. In a selective market, the firm that evaluates forty opportunities beats the firm that evaluates fifteen, almost every time.”
From a Multifamily Tool to a Firm-Wide Capability
QuickData.ai started as a focused answer to a focused problem. Multifamily underwriting depends on three core documents: the rent roll, the trailing twelve-month operating statement, and the offering memorandum. None follows a universal format. A rent roll from a 50-unit property in Dallas looks nothing like one from a 300-unit complex in Atlanta, and generic extraction tools trained on invoices or medical records cannot tell gross rent from effective rent or a capital expenditure from a recurring maintenance expense.
So Bratslavsky trained QuickData.ai specifically on multifamily documents. It runs as an Excel add-in rather than a standalone application, extracting data directly into whatever underwriting model the analyst already uses. The company reports 98% accuracy on rent-roll extraction, a 97% identification rate for T12 line items, and more than 800 multifamily professionals on the platform at $99 per user per month, with users saving an average of 15 hours a month.
But the tool, it turns out, was the entry point rather than the destination. What Bratslavsky discovered working with acquisitions teams is that document extraction is one repeatable task among dozens inside a commercial real estate firm, and the same approach that automates a rent roll can automate lease abstraction, investor reporting, expense benchmarking, or due diligence checklists across office, retail, industrial, and hospitality portfolios. The constraint was never the technology. It was that nobody inside the firm knew how to apply it.
That realization reshaped the business. Today, Bratslavsky, who has worked as a fractional chief technology officer, spends most of his time consulting with commercial real estate companies to help them learn AI and implement it in-house. He works alongside each department to co-create workflows and tools that automate repeatable tasks, and he does it in a way that teaches people to keep automating on their own after the engagement ends.
Teaching Firms to Automate Without Writing Code
The reason this model works now, when it would not have worked three years ago, is that automation no longer requires programmers. Tools like Claude Skills and others let anyone automate a workflow by properly describing their process, in plain language, so that AI can carry it out. If an analyst can explain how they benchmark operating expenses or how they build an investment committee memo, that explanation can become a reusable, automated workflow.
This is precisely the process Bratslavsky teaches his clients. Teams get comfortable with what AI can actually do for them, build their first automations with his guidance, become independent over time, and develop the habit of adapting as new AI capabilities come to market. The goal is not dependence on a consultant. It is a firm that can keep improving its own operations after he leaves the room.
Security is the other half of the engagement. Commercial real estate firms handle sensitive financial data, investor information, and confidential deal terms, and the flood of new AI products includes plenty that should never touch any of it. Bratslavsky screens tools for his clients to ensure they are secure and that data privacy is maintained before anything is deployed.
“What they do with the time is the interesting part,” he says. “Nobody goes home early. Analysts benchmark expenses against more comps, asset managers dig into the line items they used to skim, and acquisitions teams look at deals they would have passed on for lack of bandwidth. The judgment work expands to fill the space the mechanical work used to occupy.”
The broader shift is not about whether AI belongs in commercial real estate. That question has been settled by adoption. The more precise question is where AI adds value without introducing new risk. Bratslavsky’s answer is specific: automate the repeatable work, validate the output, keep the data secure, and teach the humans to do the work that requires human judgment, plus the automating itself. In a market where capital is active, pricing has stabilized, and the best opportunities go to the firms that can evaluate the most deals, that capability may be the most consequential operational investment a commercial real estate company makes this year.
