What do real estate agents want AI to do?
When we ask Tuesday masterclass attendees what they want an AI employee to handle, the people in real estate who answered mention workflows, admin, social media and marketing, and organizing and replying to email.
That matches the room overall. Out of 174 business owners polled across 19 industries, email management and replies came first, then social content, then reporting. For an agent, email is where deals slip, because a showing request that sits all afternoon can easily go to whoever answered first.
Which tasks should an agent hand to AI first?
Start with whatever you write the same way every week. Inquiry replies, listing descriptions and check-in emails to your sphere all follow a pattern you already know, which makes them good first AI employees.
Then add the connections. Once it can read your calendar it can offer open showing times. Once it can read your CRM it knows who's a past client and who's a new lead. Once it runs on a schedule, the deadline list is waiting for you before your first coffee.
How does an agent keep the personal touch?
Everything client-facing stays a draft. The AI employee writes, you approve, and you send. That's the same idea as the confidence rule in Luke's AI bookkeeper: when it's 90% sure or more it does the work, and anything under that gets flagged for a person. On its last run it coded 187 transactions and flagged 14.
On contracts, black out client names and personal details before you upload a document. Claude can still read the terms, dates and contingencies without knowing who the buyer is.
If you want this built inside your business with you instead of figuring it out alone, that's what the AI Operating System is for.