Most consultants are hired to produce something and move on, leaving the client to hire them again later. David Bratslavsky takes his practice in the opposite direction. He finds his practice inside a commercial real estate firm, sees the work that people there do week by week, automates it with them, and teaches them to build the next automations without him.
Bratslavsky is the founder of QuickData.ai and a multifamily investor who has underwritten his own deals and has spent more than a decade building software, including as a CTO. His argument is that the AI producing durable value in his industry is not the kind shown at a keynote. It is operational, specific, and only obvious to someone who has done the job.
The hype version of AI in real estate is computer vision scanning buildings. The working version is an analyst who no longer retypes a rent roll, an asset manager whose monthly reports enter themselves, and a group that knows exactly how both of those things happen.
What a Firm Gets From the Engagement
The first benefit is time. One acquisitions team he worked with was saving 15 hours per person every month after rent rolls and T12 statements went straight into the firm’s own Excel underwriting model. An asset management group that had been typing monthly reports from 26 properties into a database by hand now has them entered automatically, with the property-to-property comparison ready when the reports arrive. A brokerage produces the rent roll, T12, and unit mix summaries for each offering memorandum without rebuilding them. A property operations team has a chatbot answering applicant and resident questions at any hour.
The second benefit is that nothing is a black box. Most of these automations are written in plain English via tools like Claude and ChatGPT, so the people who do the work can read and change them, and write new ones. No one needs to know how to code.
The third is judgment about what not to automate. Assumptions, pricing, relationships, and the decision to buy stay with people. Bratslavsky also helps firms, task by task, decide whether to buy a tool off the shelf or build the workflow in-house, so they don’t pay for software they could have built or build something they could have bought.
How the Work Is Done
The practice serves investors, lenders, and brokers, along with the property operations teams that support them. The method is consistent. Bratslavsky learns how the company does its work today, department by department. He identifies the repetitive tasks worth handing over. He and the team build the automation inside the tools they already use and test it against real cases. The employees who own the task take part in building it, so they understand what the AI is doing and how to maintain it.
As the firm’s own description of its work is, he “trains each department to create and upgrade their own automations, so they can keep adapting as the business changes without calling in a consultant every time.”
That is the part that matters most. The model is designed to make the consultant unnecessary. Once a team has built one working automation and understands the process, they can apply the same method to the next task. The firm keeps the capability instead of outsourcing it permanently.
Where It Started
The consulting grew out of a product. Bratslavsky first built QuickData.ai, a tool that extracts data from multifamily rent rolls and T12 operating statements and places it in an Excel underwriting model. It reads Excel, CSV, and PDF files from Yardi, RealPage, Entrata, AppFolio, and other property management systems, returns one row per unit or one categorized line per account, and checks every result against the totals the document itself prints. The company reports 99.3% accuracy. It is sold as an Excel add-in at $99 per person per month, with an API for developers, and the first 20 documents are free.
The tool does one thing, inside the application its users already have open. Building it taught Bratslavsky that every firm has dozens of similar tasks in plain sight: clear inputs, a known correct output, and a person doing the mechanical part by hand.
Why the Big Platforms Keep Failing
The pattern in real estate technology is familiar. A vendor comes in with a platform that promises to handle everything. Implementation takes months. By the time it is live, workflows have changed, staff resent the new interface, and analysts quietly return to their spreadsheets.
Bratslavsky’s approach is the opposite. Start with one specific task. Build inside the tools the team already uses. Let the people who do the work help build it. Expand only after the first automation runs reliably.
It will not generate conference invitations. It may save a firm a thousand hours a year while the flashier tools are still in the demo phase. More to the point, when the engagement ends, the firm is left with people who know how to keep going.
