Welcome to the CRE AI Studio newsletter. Your regular dose of AI news and tips on putting it to work in commercial real estate.
OpenAI launches ChatGPT Work, deepening race for workplace AI tools ChatGPT Work (launched July 9) is an agent that pulls context from your connected apps and files, then independently produces finished documents, spreadsheets, decks, and reports over multi-hour sessions. Rolling out now on Pro, Enterprise, and Plus.
Microsoft 365 Copilot Gets 40+ Updates in July: Claude Support, New Interface Microsoft shipped 40+ Copilot updates this month, headlined by Anthropic's Claude as a selectable model inside Copilot Chat and smarter document navigation that cites specific sections.
Redeploying Claude Fable 5 Anthropic's frontier model Claude Fable 5 is back for everyone after US export controls were lifted July 1 — and free access on Pro, Max, Team, and premium Enterprise plans was just extended a third time through July 19.
Claude Cowork expands to mobile and web Anthropic's Cowork agent — previously desktop-only — rolled out to claude.ai and iOS/Android in beta on July 7, starting with Max subscribers, with doubled usage limits through August 5.
We’re Launching Something New: The Claude Cohort for Commercial Real Estate ⚡⚡
|
Start Date: August 13th, 2026
Format: 6 weekly live lessons, 1 hour each
Level: All levels welcome
Includes: Live + recorded sessions, 3 months free access to the CRE AI Studio
Your last LinkedIn post is a lead list.
The people who comment on your posts are your warmest leads. They know you, they engaged, and they told you what they care about in the comment they left. Then they sat in your notification feed until they disappeared.
This week's lesson fixes that. Grab the URL of a post that landed with your audience and hand it to Claude. It pulls every commenter through Apify: name, comment, headline, profile URL. Then it runs the list through Apollo for verified emails, matching in batches and counting credits as it goes. Names that come back masked get a recovery pass instead of getting dropped. Export to a spreadsheet. No code.
On my run: 78 commenters, 42 matched, 37 verified work emails, 42 Apollo credits. It also flagged a commenter who runs sales at an AI-for-CRE startup — a partner conversation I would have scrolled past.
Credit to Pietro Montaldo for this one. The full build, with the exact prompts and both connectors, is in the lesson: Week 33: Finding More Leads with Claude Cowork using Apify and Apollo
Don't stop at LinkedIn comments
The comments workflow is one recipe. The two connectors underneath it cover a lot more ground.
Apify is a marketplace of pre-built scrapers you point at almost any site. For CRE, that means:
Comp intel: scrape Apartments.com, Zillow, or LoopNet for asking rents, concessions, unit mix, and days on market in any submarket
Competitor watch: pull Google Maps reviews and ratings on the properties you compete against
Development pipeline: set a recurring scrape on permit and local government sites to catch new supply before it hits the market
Prospect lists: company directories, conference attendee lists, any public site with names on it
For deal sourcing instead of people sourcing, the same setup runs our prospecting build: Turn Claude CoWork Into Your Prospecting Pipeline.
Apollo is a B2B database of hundreds of millions of people, plus the tools to reach them:
Search by title, market, and company size to build target lists: owners, principals, LPs, brokers at specific firms
Drop in a name, company, or LinkedIn URL and get back a verified email and direct dial
Bulk-enrich a stale contact list or CRM export before a campaign instead of guessing
Watch job changes and new postings, then reach out when a firm is growing, not after
The part most people skip is scheduling. A one-time scrape is a report. A scheduled scrape is a market intel feed that updates itself.
And when the target is a property, not a person: How to Find Out Who Owns Commercial Real Estate with AI.
Pull your best post from the last 90 days and build your list.