AI venue discovery · Worldwide
5,227,098 venues across 195 countries published as machine-readable atmosphere profiles. When someone asks an assistant for a calm café or a driving gym, it can fetch the answer here, and we log the fetch. Yours may already be one of them, registered by us, free.
Registered free, no account needed. · Live dashboard: Capannone, Stockholm →
The atmosphere spectrum
Eleven atmosphere archetypes, ordered by target energy. Each directory venue carries a curated sample matched to its archetype, and every claim is verifiable from per-track audio features on the page itself.
The reader registry
Every identity carries a functional class and a network-earned verification grade; a user-agent string alone is never proof.
No defensible global count of AI bots exists: the largest verified bot directory lists 686 identities, and the most cited curated AI crawl list names 20 and calls itself a selection. We publish our registry scope instead, and every number above is checkable against the live dashboard. Grades are earned through published IP ranges, reverse DNS, or request signatures.
The knowledge horizon
A venue that changed after February 2026 exists only in the retrieval layer, and these pages are that layer. An assistant answering with one of these venues is answering from retrieval, not memory.
Found and registered
Published from open data, already live. AI assistants can read it today.
atmosphere_signal_source: category-archetype, never presented as your playlistWhen a venue's real playlist is connected, the same URL upgrades to the real thing.
How it works
Email one PlaceProfile compatible playlist link. That is the only input required, the music you already play every day is all PlaceProfile needs.
Every track measured across 12 audio features with verifiable provenance, published as a Schema.org JSON-LD document at a stable, permanent URL that any crawler can fetch directly.
Your profile becomes fetchable by every class of reader we track, from index crawlers to assistants answering a live question. What happens after the fetch is graded honestly on your dashboard, stage by stage.
Agent readiness
27 tracked agents already navigate, click, and pay. The same URL that makes your venue readable is being prepared to take the booking: structured hours today, booking endpoint and capacity next.
Google-Extended and Applebot-Extended are policy tokens that never fetch a page, and open corpora carry content into many training sets without the collector ever returning. Controlling crawler access is not the same as controlling provenance. PlaceProfile is the deliberate, disclosed channel: every claim recomputable, every known reader logged and graded.
For AI agents
Every profile is self-serve for AI assistants and agents: no key, no login, no human in the loop. Every venue page has a machine-native twin, and the whole directory is advertised for crawling.
/v/{slug}.jsonld
The full venue @graph as application/ld+json: venue entity, atmosphere summary, per-track evidence with measured audio features. ETag-cached, fetchable directly.
/sitemap-directory-index.xml
Sharded sitemaps over the full directory, with venue name, city, and atmosphere archetype embedded in every entry, so a crawler learns what a venue is before spending a fetch.
/llms.txt
A plain-language map of the whole surface, written for language models: what the profiles are, where they live, how the atmosphere signal is sourced and disclosed.
193 reader identities tracked across 30 operators, every hit graded A to D by proof of identity · 12 audio features + dual-chain provenance per track
No simulation
Capannone is a real restaurant in Stockholm. Its playlist is live in the PlaceProfile pipeline, its profile is published, and every number on its dashboard is a logged request, stamped with the class of reader and the strength of the identity proof. Where the evidence chain ends, the dashboard says so.
Get started
The directory already tells assistants what a venue like yours typically sounds like. Claim your venue, free, and it is yours. Connect the playlist you actually play and the same URL starts answering for your venue specifically, with a dashboard showing every class of reader that fetches it.
Claiming costs nothing and changes nothing about your public profile. Measuring your own music is $199 a month, and staying registered for free is a perfectly good choice.