A Multifamily Broker Went From 200 Listings a Week to 15–20 Matched Deals Each Morning
An apartment-building broker built an AI system that screens new listings against buyers' criteria and drafts first-pass pro formas in about 10 minutes instead of 2–3 hours, and says his active listings grew from 5 to 20.
Who this is about: Multifamily (apartment) real estate brokerage
What it takes
- Cost
- Cost not statedHe built his own system and doesn't name the tools.
- Setup time
- A few weeksOur estimate from the source, for one business.
- Skill needed
- Developer skillsYou build or run your own code, even if AI writes it.
Explain it simply
The broker used to skim hundreds of apartment listings by hand. His AI system now scans new listings, reads rent sheets (turning photos of pages into text), and emails him only the ones that match what his buyers want. It also drafts a first money model for each pick so he spends his time judging deals, not sorting paperwork.
What was done
After six months of testing, the broker uses AI for the admin around deals. A system scans new listings, reads rent rolls with OCR and flags properties that match his buyers' criteria. For underwriting, AI pulls comparable sales and rent trends and builds an initial pro forma. He keeps real comp data in a Google Sheet he fills in by hand, which feeds the system.
How it went
He went from manually reviewing 200+ listings a week to 15–20 matched deals emailed each morning, and found 3 deals he says he would have missed. First-pass pro formas take about 10 minutes instead of 2–3 hours, and get him about 80% of the way. He credits the freed time for growing from 5 to 20 listings.
- Limits: AI valuations were a "complete disaster" because the model misses local details such as building problems or how a new transit line affects a neighborhood, so he still does valuations himself. Every pro forma needs checking. These are his own figures from a Reddit post.
Prompt
Each morning, scan new listings from [sources]. OCR any rent roll and extract
units, rents and occupancy. Compare each property with these buyer criteria:
[market, unit count, price range, cap rate, condition]. Email me the 15-20
best matches with a one-paragraph summary each.
For any property I select, draft a first-pass pro forma using the verified
comps in [Google Sheet]. Mark every assumption you made. Do not estimate
market value.Our version; no exact prompt was shared.
Business fit
Who this helps, and how:
Local service businesses: screen incoming job requests or bids against your criteria so you only review the good fits.
Local & serviceStartup founders and e-commerce sellers: scan supplier or marketplace listings daily and get a short list that matches your buying rules.
Shopify & ecommerceIndependent contractors (barbers, bartenders, consultants): consultants who analyze deals can draft the first model in minutes and spend time on judgment.
AgenciesAgencies and freelancers: build a daily "matched opportunities" email for clients who buy property, equipment or businesses.
AgenciesMechanics and tradespeople: contractors bidding on jobs can pre-screen posted projects by size, location and scope.
Local & service
Related
- Evidence grade CA 70-Year-Old Manufacturer Cut RFQ Turnaround Nearly in Half With Claude CodeVendor or secondhand, using Claude Code
- Evidence grade CA Nine-Person Architecture Practice Sends Fee Proposals a Day After the Site Visit, Not TenVendor or secondhand, using Custom agents built by Blott
- Evidence grade CA Travel Advisor Pointed Codex Desktop at Client Notes and Got a Priced Hotel ShortlistVendor or secondhand, using Codex
- AI in sales examplesDepartment hub: AI for sales, 34 examples
Similar tools
Other tools for this job, and how they differ. Chosen by our editors, never paid for.
- ChatGPT
Can draft first-pass summaries and pro forma outlines from deal notes you paste in; it does not scan listings for you by itself.
- Claude
Useful for careful underwriting write-ups once you already have the matched deals and comps.
- CompStak
A commercial real-estate data product for verified comps, not a DIY listing scanner.
“Strong evidence” (grade B) means: The person who did the work describes the setup and results in detail. How we grade. Spotted a mistake? Suggest a correction.