Skip to content
Start a project

Real estate · AI services

Home search by plain-language question, with an honest 'nothing fits'

Buyers ask 'anything in Austin under 1.2m?' in chat and get a link to a search form.

An assistant that reads a budget and a market from a plain question, shows up to three listings, and says so when nothing fits.

Try it on the live demoBuild something like it

This is a concept build. Everline Realty is a fictional business. The code is real: open the demo and try the steps below.

The everyday problem

Buyers ask 'anything in Austin under 1.2m?' in chat and get a link to a search form.

The chat reads the question and shows the listings that fit.

The hard part

  1. A budget typed loosely

    'under 1.2m', 'below 750k' and 'under 2' are all read as a dollar cap.

  2. A neighborhood, not a city

    Coral Gables or Paradise Valley match their market.

  3. A sold home offered

    Sold listings are left out before matching.

  4. Nothing in budget

    It says so and offers an off-market search. No listing is invented.

How it runs

Workflow

How a plain question becomes listings, or an honest no.

Sample run
  1. Step 1: Question.
  2. Step 2: Read the budget. "under 1.2m", "below 750k".
  3. Step 3: Match the market. City or neighborhood across six markets.
    If this fails: No city or budget: knowledge-base answer instead.
  4. Step 4: Skip sold listings.
  5. Step 5: Up to three listings.
    If this fails: Nothing in budget: says so, offers off-market search.
  6. Step 6: A person on request. Asking for a person gets the hand-off.

Tap “On fail” to see what happens when a step breaks.

Edge cases we handled

What the system does in the unusual cases. Where there is a phrase, type it into the demo.
WhenWhat the system doesTry it
A buyer asks for 'homes in Austin under 1.2m'.Up to three listing cards in budget.homes in Austin under 1.2m
A buyer asks for 'anything in Austin under 100k'.An honest 'nothing fits', and an offer to search off-market.anything in Austin under 100k
A budget is typed as 'under 2'.Read as $2 million.No phrase to try
A buyer asks to talk to a person.The hand-off message, with a booking option.talk to a person

Guardrails and hand-off

Guardrails

  • Sold listings are never shown.
  • It shows only listings in its own data.
  • A person is one message away.
  • The contact server has a honeypot and a rate limit; the live demo is static, so it is not running.

When a person takes over

Asking for a person ('talk to a person', 'real person', 'call me', 'agent now') gets the hand-off message and a booking option.

Not in this build yet

We would rather you hear it from us.

  • The chat runs in demo mode with no server.
  • The live demo is static: the contact server, with its honeypot and rate limit, is not running.
  • Listings and agents are composites.
  • Sending offers and contracts to a person is a written rule in its knowledge base, not code.

Try it yourself

  1. Type: homes in Austin under 1.2m

    Expect: Up to three listing cards.

  2. Type: anything in Austin under 100k

    Expect: An honest 'nothing fits', no invented listing.

Open the chat on the demo site. Please don't type real personal details.

Built with

  • React
  • Node.js

  • No. In this concept build, the chat runs in the browser with no server. In your build, hand-offs reach your team through the tool you already use.

  • Yes. It reads your own listings and knowledge file, so your markets, prices and answers drive it.

More work

How Copperline triage chat runs. Steps: Message in chat, then Safety check on every turn, then Read the problem, then Check the area, then Pull out contact details, then Offer a time, then Office board.

  • Concept build
  • Plumbing
  • AI services

Plumbing chat that puts safety before sales

Everyday: Panicked messages arrive at night. The chat either says nothing useful or pushes a booking form.

The hard part: A gas smell mid-conversation

Read the build: Plumbing chat that puts safety before sales Open demo of Copperline triage chat (opens in a new tab)

How Kestrel and Dune Lake triage runs. Steps: Request arrives, then Read urgency, then Emergency?, then Pick the team member, then Hold the slot, then Office board.

  • Concept build
  • Property management
  • AI automation

Maintenance emergencies triaged in each company's own words

Everyday: Resident maintenance requests, owner questions and viewing requests all land in one place, and a routine request looks the same as an urgent one.

The hard part: Emergencies that differ by business

Read the build: Maintenance emergencies triaged in each company's own words Open demo of Kestrel and Dune Lake triage (opens in a new tab)

How CrestBay answer desk runs. Steps: Question, then Hand-off words, then n8n workflow, then Server knowledge base, then In-browser retriever, then Logged with its source.

  • Concept build
  • Roofing
  • AI services

A roofing assistant that keeps answering when the AI goes quiet

Everyday: The same few dozen questions about pricing, repairs and insurance arrive every day, and the answers change with who picks up.

The hard part: The AI service going quiet

Read the build: A roofing assistant that keeps answering when the AI goes quiet Open demo of CrestBay answer desk (opens in a new tab)

Start a project

Want something like this for your business?

A 30-minute call, then a written plan and a fixed quote. No obligation.

Book a workflow auditSee the systems
  • A reply from a person within one business day.
  • Fixed price in writing
  • You own everything