Brickfold

Calgary · live listing data · AI search

Find real estate by asking.

Describe the building, parcel, neighbourhood or opportunity you want. A language model reads the sentence, works out whether you are renting or buying, and turns it into a real structured query across live Calgary listings.

587

Active rentals mapped

1,797

Apartments for sale

312

From individual landlords

3

Sources normalised

Two sides of the same market

Ask a question and Brickfold sends you to the right one — or go straight in and use the filters.

Rentals

Calgary rentals map

Every active rental we track from RentFaster, Kijiji and Facebook Marketplace, normalised onto one map with the individual landlords marked.

Open

For sale

MLS apartments

Active apartment and condo listings with price per square foot, condo fees, age, exposure and days on market — and a pane you can simply talk to instead of setting thirty filters.

Open

The search layer

A language model sits between your sentence and the database.

Not a chatbot bolted onto a listings site. The model's only job is translation — from how people describe a home into the structured query the data can actually answer.

  1. 1

    Reads the question

    You write a sentence, not a filter combination. Brickfold shows its reading of it — every criterion it understood and every word it could not use — before you commit to a search.

  2. 2

    Picks the right market

    “2 bed under 2000” and “2 bed under 400k” differ by one character and mean opposite things. A model settles rent versus purchase from the wording and the scale of the number, and tells you which it chose.

  3. 3

    Becomes a real query

    On the apartment map the sentence is turned into the complete filter state — price, beds, baths, communities, condo fee, tax, floor, exposure, square footage, year built, days on market, amenities — and the map redraws from it.

Answers come from the data

The model never sees a listing and never writes a line of SQL. It returns one object describing the search; your database returns the properties. So the prices, counts and addresses on the map are the real ones — there is nothing for it to invent, because it is never the thing answering.

It shows its work

Every search tells you what it understood, what it had to ignore, and which market it sent you to. The filter panel stays open beside the map with the values the model chose, so you can correct any one of them by hand and keep going.

Who it is for

Investors
Find buildings by unit count, age, location, rents, ownership or redevelopment potential.
Brokers
Turn client briefs into fast property shortlists and comparable searches.
Developers
Explore parcels, zoning, demographics, supply, demand and nearby infrastructure.
Renters & buyers
Describe the lifestyle and constraints that matter instead of configuring dozens of filters.
Data teams
Expose structured property databases through a controlled natural-language query layer.

Real estate, unfolded.

Talk to the market, not the filter panel.

Open the map