David Bratslavsky on Cross-Border Real Estate: Highlights From ReShaped 2025
A recap of David Bratslavsky's ReShaped 2025 panel on AI's role in global real estate investment and management.
Read post
Back to all posts
Speaking
David Bratslavsky at CRE AI Studio: A Live Demo of AI-Driven Underwriting
David Bratslavsky at CRE AI Studio: A Live Demo of AI-Driven Underwriting
David Bratslavsky took the stage at CRE AI Studio with a single goal: skip the slides and show the audience what AI-driven underwriting actually looks like in real time. What followed was a live demo of QuickData.AI running through a real rent roll, a real T12, and a real Offering Memorandum — start to finish — in front of a roomful of skeptical CRE professionals.
This post recaps that demo, the audience reaction, and the questions David Bratslavsky kept getting asked afterward.
The 12-minute demo
The setup was deliberately mundane. A 180-unit garden-style rent roll PDF, a one-year T12, and a 60-page OM — the kind of file packet that hits every multifamily analyst's inbox three times a week. David Bratslavsky timed himself.
- 0:00 — Drop three files into QuickData.AI.
- 0:45 — Rent roll parsed. Unit-level data extracted at 98% confidence.
- 1:30 — T12 categorized against the firm's chart of accounts.
- 2:10 — OM rent comps and pro forma extracted.
- 2:30 — Excel model populated. Underwriting can begin.
Two and a half minutes of automated work to replace what most teams call a "Saturday afternoon."
"The point of the demo isn't speed. It's reliability. The same workflow runs the same way 100 deals from now." — David Bratslavsky
Audience questions that mattered
After the demo, David Bratslavsky took 20 minutes of Q&A. A few questions came up repeatedly.
"What about edge cases?"
David Bratslavsky's answer: the platform doesn't pretend to be perfect. Confidence scores flag exceptions. Humans review them. The win isn't 100% accuracy — it's exception-based review instead of full re-typing.
"Does it work on hand-written rent rolls?"
Yes, with caveats. Hand-written documents flow through the same pipeline but at lower confidence. They surface more exceptions for human review. The platform never asks the user to trust output it isn't sure about.
"What's the security model?"
QuickData.AI runs on enterprise-grade infrastructure with full audit logging. Documents are processed in-region and retention is configurable. David Bratslavsky walked through the SOC 2 path the team is on.
"Can it integrate with our existing model?"
This is the question David Bratslavsky lives for. The answer is yes — and that integration is the whole point. The platform writes to whatever Excel template the firm already uses, no template migration required.
What the demo proved
CRE AI Studio attendees came in with the right skepticism: AI demos are often canned, scripted, and unreliable in the wild. David Bratslavsky used three real, unedited files chosen by the moderator, and the platform handled all three. The room shifted. The conversation moved from "is this real?" to "how do we get on this?"
"If your demo can survive a moderator picking the files, you have a product. If it can't, you have a deck." — David Bratslavsky
Next stops
The CRE AI Studio appearance was one of several speaking engagements David Bratslavsky has lined up over the next few quarters. If you're attending an industry event and want a live demo, reach out — he's an easy person to get on a stage.
About the author: David Bratslavsky is the founder of QuickData.AI, a fractional CTO, and a Member of the Forbes Technology Council. Connect with him on LinkedIn.