A website chatbot for real estate agents and brokerages

A buyer scrolling your listings at 10 PM has a specific question about a specific house: "Is the basement finished?", "What are the HOA dues?", "Can I see it Saturday?" An Allinaa assistant answers from that listing's details, asks only the qualifying questions that are still missing, and hands your agents a showing request with the buyer's permission to call.

Listing-page questions answered from that listing's details

The assistant knows which page the visitor is viewing and reads its visible text: price, beds and baths, square footage, the description, and the details your website shows. That means "does this one have a garage?" gets an answer about this house, not a generic reply. Two things make it work well:

• Listing pages with real text. If key facts appear only in photos or a virtual tour, the assistant cannot read them.
• Honest gaps. If the listing does not say whether the basement is finished, the assistant should say that the listing doesn't mention it and offer to have the agent find out. It answers only from your information and the page, and it removes any link that isn't on the page or in your approved knowledge.

Listing data that comes into your site through a feed changes often. The assistant reads the page as it is when the visitor asks, so a sold or price-reduced listing is reflected as soon as your page updates. More on page-aware answers.

Buyer or seller? Qualifying with up to three questions

Nobody wants to fill out a twelve-field form to ask about a house. Allinaa's assistant asks up to three short clarifying questions, one at a time, and skips any the visitor has already answered. For real estate, the useful ones are:

• Buyers: timeline (this month, this season, just looking), whether they are pre-approved or paying cash, and the areas they are considering
• Sellers: the property address or neighborhood, timeline, and whether they are also buying

Then it offers the inquiry form. The visitor confirms their contact details and agrees to be contacted, you get an e-mail notification, and they get a confirmation e-mail.

Showing and valuation requests your agents follow up

Allinaa does not connect to MLS systems, showing services or your CRM, and it does not book showings. It collects the request (which property, preferred times, contact details) and your agent confirms. For sellers, it collects a valuation request; it never estimates a home's value itself. Requests wait in your inquiry inbox with a status, private notes, search and CSV export, which you can import into whatever system your brokerage uses. See lead capture.

Fair housing: property facts, not neighborhood steering

A chatbot on a real estate website must follow the same rules as your agents. The federal Fair Housing Act prohibits discrimination based on protected characteristics such as race, religion, national origin, familial status and disability; see HUD's overview of the Fair Housing Act. Questions like "is this a safe neighborhood?", "what kind of people live here?" or "are the schools good?" can lead to steering if answered with opinions. Write a standard response into your knowledge that points to objective sources (for example, the school district's own website or local police crime maps) instead of characterizing a neighborhood. Then test the assistant with those questions before you go live, and ask your broker or compliance contact to review the wording. AI answers can be wrong, and you remain responsible for what your website says.

International and non-English-speaking buyers

If you work with relocating families or international buyers, language matters. Standard includes 3 languages: American English plus two you choose, often Spanish and one more. Premium covers all supported languages, 108 in the chatbot catalog today. The listing stays in English; the assistant explains it in the buyer's language. For contracts and disclosures, buyers should still rely on your agents and qualified interpreters. See multilingual support.

Leads that arrive after hours and on weekends

Buyers browse in the evening and on Sunday mornings, when agents are at open houses or with family. The assistant answers listing questions at any hour and captures the request. Your agent sees the details first thing and calls back with context: which house, what they asked, and when they want to see it. Conversations also continue across pages and return visits, so a buyer who looks at three listings sees one continuous chat.

Illustrative conversation

Illustrative example, not a real conversation.

If you also manage rentals, see property management.

Questions owners ask

Does the chatbot connect to the MLS or my CRM?

No. It reads the listing page the visitor is viewing and your approved knowledge. Inquiries can be exported as CSV for your own systems.

Can it book showings?

It collects showing requests with preferred times. Your agent confirms the appointment.

How does it handle questions about neighborhoods and schools?

With the response you write: point to objective sources and avoid characterizing neighborhoods. Have your broker review that wording and test it before launch.

Will it estimate what my home is worth?

No. It collects a valuation request so an agent can prepare a market analysis.