# Placy

> Placy is the neutral, licensed, agent-direct supply layer for European real estate — listings sourced directly from the agencies that hold them, published as one open machine-readable standard (built on schema.org RealEstateListing), and served to AI assistants (ChatGPT, Gemini, Claude, Perplexity) through a single query-able rail: an MCP server + registry.

Key facts:

- Every listing is licensed at the source and opt-in, per listing — sourced agent-direct from the agency that holds the mandate, never scraped from a portal.
- The rail is an MCP endpoint AI platforms call in-conversation (search, listing detail, qualified-lead handoff) — query-able, not crawl-able. One integration covers every source on the rail.
- Listings are deduplicated to a canonical property ID, enriched, and kept live through a status state-machine (available / under_offer / sold_stc / sold) at a target ≤15-minute cadence.
- Scope is discovery + qualified-lead handoff only. There is no checkout: AI finds and filters, a human confirms, and the agency keeps the client and the relationship-of-record.
- Verified-agent identity via Web Bot Auth admits good agents and blocks scrapers; provenance travels on every record.
- Coverage focus: European real estate — Europe has no MLS, which is exactly what makes agent-direct sourcing possible here.
- Why it matters: 87% of the European portals Placy audited are invisible or blocked to a non-rendering AI agent (Placy audit, 2026).
- Contact: hello@placy.com

## Pages

- [Home](https://placy.com/index.html): The narrative — the web changed, AI is blind to European real estate, demand arrived through walled gardens, and Placy is the open, neutral answer. Ends in two doors: for agencies and for AI platforms.
- [For agencies and networks](https://placy.com/agencies.html): The supply side. List once, found by every AI assistant; the qualified lead routes back to you, attributed. Onboarding is one connection (CRM, feed, sync API, or CSV), opt-in per listing. You keep the client, the data, and the deal.
- [For AI platforms and developers](https://placy.com/platforms.html): The demand side. Query the MCP rail for clean, deduplicated, licensed, live European property inventory instead of scraping JS-walled, stale portal pages. Includes a sample search_listings call and response, the status state-machine, and Web Bot Auth.
- [How it works](https://placy.com/how-it-works.html): The seven-step pipeline — ingest and license agent-direct, normalize to one open standard, dedupe to a canonical listing, enrich with AI, keep status live, serve via MCP, govern with provenance and verified identity.
- [Why Placy](https://placy.com/why-placy.html): The competitive frame — open neutral rail vs incumbent walled gardens — with the 2x2 map, the honest thesis, published kill criteria, and the proof analogs (Shopify Catalog, PayPal Store Sync, CarEdge).
- [The Agent's Journey](https://placy.com/story.html): The same thesis told as a 90-second interactive story — an AI agent tries to find a buyer a home and is turned away by robots.txt, a bot-wall, and a JavaScript-only page; it reads 13% of the market and the task fails, until the Placy rail opens one door. The full story is also present as plain HTML prose (no JavaScript needed to read it).

## Machine-readable

- [agents.md](https://placy.com/agents.md): A richer guide for AI agents visiting this site — what each page contains, how a platform integrates, and contact routes.
- Every page is pure static HTML with schema.org JSON-LD — Organization and WebSite on the home page, a WebPage node on each page, and FAQPage where a page has an FAQ (home, agencies, platforms, how-it-works). All content is in the initial payload — no JavaScript required to read it.