The buyer who wanted to view the listing at 9pm didn't leave a voicemail. She just closed the tab.

An AI receptionist for real estate answers the questions buyers and renters ask about a listing — price, area, availability for a viewing, what's included — the way an agent would on the phone, but on the agency's website, at any hour. Maya reads the agency's real listings and pages to configure itself, so a visitor browsing at 9pm can ask about a specific property in their own language, get a grounded answer, and book a viewing or leave qualified details instead of bouncing to the next listing site.

Last updated 2026-07-16

Where a phone-based agent stops

An estate agent answers calls during working hours, one conversation at a time. But a large share of property browsing now happens on the website, in the evening, on a phone, from someone who isn't ready to call a stranger yet — they'd rather look at one more listing than dial a number. That visitor doesn't leave a voicemail when a question goes unanswered; they leave the site, and the agency never even learns they were interested.

Maya covers that door. Once an assistant can answer a listing question from the agency's real data, qualify what the visitor actually wants (budget, area, timeline) as they say it, drive them hands-free to the right listing, and book the viewing — "receptionist" is the small version of what it does. A missed call just sits there, unanswered, until someone happens to check it; Maya is closer to someone who actually picks up and sorts the enquiry out on the spot.

What Maya does for a real estate agency

It sets itself up from your site — you don't build it, you point it at your listings — Maya reads the agency's property listings, areas covered and pages and configures itself — no FAQ to write, no flows to draw. The agency reviews what it learned and corrects anything in a sentence, live in about ten minutes.

It greets every visitor by voice or text — in their own language, from the first word — the visitor can talk to it or type to it, and Maya opens in their language, not the agency's. For an international buyer browsing at odd hours, that first moment — answered in their own words — is what keeps them on the site instead of the next listing portal.

It answers property and area questions from your real listings — not a guess — "is this still available," "what's the commute to the centre," "does it have parking" — Maya answers from what the agency actually has listed, so the visitor gets a correct answer at 9pm instead of an email reply the next afternoon, by which time they've often moved on.

It qualifies the enquiry as it happens — budget, area, and timeline become real meaning — instead of a name and a generic "interested in this property" message, the agency gets: looking in this area, this budget range, wants to move within three months, asked about two specific listings, wants a viewing this weekend. An agent can act on that in one read instead of chasing details in a follow-up call.

It drives the site for them — scrolling and navigating to the right listing hands-free — the visitor says what they're looking for and Maya moves through the listings for them, hands-free — no mouse, keyboard, or screen — right up until they choose to book a viewing or submit an enquiry. Someone browsing listings one-handed on a phone still gets to the right property.

It follows the visitor's language — even when they switch mid-conversation — Maya replies in the language of every message, and if a visitor slips from English into Romanian mid-sentence, it follows without missing a beat or asking them to start over. When the agency's site already has that language, it switches to the correct version, so a relocating buyer can browse and ask in their own words.

It books the viewing inside the conversation — "call the office to arrange a viewing" becomes a booked slot there and then, while the visitor is still interested — instead of a callback request that competes with everything else on tomorrow's to-do list. (Booking plan and up.)

It captures the enquiry the moment the visitor would have left without asking — the browser comparing five listings at once leaves contact details and what they're actually looking for, instead of quietly moving to a competitor's site — a qualified lead instead of anonymous traffic.

What happens when Maya doesn't know the answer?

This is the question every agency asks about AI, and it's where Maya is strongest. Maya checks its answers against the agency's real listings and published details before it responds. When it can't verify something — whether a specific property is still on the market, a detail that isn't published — it says so instead of guessing, offers the visitor the right next step, and flags the question for the agency. The team can mark it handled, or teach Maya the answer in one sentence.

What that means for you: Maya gets more useful over time without ever being rewarded for confidently inventing whether a property is still available or what it includes — so a buyer doesn't book a viewing for a listing that already sold. In real estate, a wrong answer wastes a viewing slot and a buyer's trust; abstention is what keeps that from happening.