Dataset Reference
The room-level data layer missing from hotel search, AI, robotics, and asset diligence.
Roomza structures what is usually hidden behind the door: condition, layout, noise, view, amenities, accessibility, workability, sentiment, media, provenance, and refresh history.
Coverage
6,000+
Hotels in the network
Room-level
Per-record granularity
US-focused
Current coverage
Expanding
International
Current coverage is strongest in the United States, with international markets added through partner integrations, public mapping, and first-party capture.
Core vs Full
| Dataset layer | Included fields |
|---|---|
| Core | condition, layout, noise profile, view, workability, accessibility, amenities, guest sentiment |
| Full | Core plus licensed images, video, spatial data, and first-party capture assets where available. |
| Provenance | source path, collection date, rights basis, refresh history |
| Delivery | JSON, CSV, Parquet, scheduled feeds, REST API |
Field definitions
| Field | Meaning |
|---|---|
| condition | Signals tied to room upkeep, age, maintenance, and guest-facing quality |
| layout | Room configuration, sleeping setup, bathroom relationship, desk and seating placement |
| noise_profile | Known or inferred noise exposure from street, elevator, hallway, mechanical, or venue sources |
| view | Exterior or interior view type, exposure, and quality signal. |
| workability | Desk, seating, outlets, lighting, WiFi signal, and comfort for laptop use. |
| accessibility | Mobility-relevant features, barriers, and verified accessibility flags |
| amenities | In-room features, equipment, bath, bedding, minibar, appliance, and comfort items |
| guest_sentiment | Structured themes from guest feedback tied to room attributes |
| images | Licensed room-level images where available |
| video | Licensed walkthrough or room-level video where available |
| spatial_data | Room geometry, spatial captures, floor plan signals, or fixed-object placement where available |
Sources
- +Direct hotel partnerships
- +PMS integrations, including Mews and Opera Cloud where available
- +First-party Roomza Vision capture from travelers using the Roomza consumer app at roomza.com
- +Rights-reviewed structured mapping from publicly available hotel, room, and review sources.
Provenance
Every record in the corpus includes provenance metadata such as source path, collection date, rights basis, and refresh history. For commercial buyers, Roomza can provide a provenance review as part of diligence and licensing.
Where data is sourced from third parties, Roomza maintains the rights required for the licensed use case.
Sample record
Per-room record
JSON{
"hotel_id": "rz_h_4827",
"room_id": "rz_h_4827_1408",
"market": "New York, NY",
"room_type": "King Deluxe",
"condition_score": 8.1,
"noise_profile": {
"street": "moderate",
"hallway": "low",
"elevator": "none_detected"
},
"view": "city",
"workability_score": 8.7,
"accessibility_flags": ["step_free_route"],
"amenities": ["desk", "mini_fridge", "blackout_shades"],
"guest_sentiment_summary": "liked for quiet, desk setup, and light; mixed comments on bathroom age",
"source_path": "direct_hotel_partner",
"collection_date": "2026-05-01",
"rights_basis": "commercial_license",
"refresh_history": {
"last_refreshed": "2026-05-15",
"refresh_cadence": "monthly"
}
}Quality controls
Source tagging
Every record is tied to its source path and collection method.
Normalization
Room fields are structured into consistent attributes across brands, markets, and source types.
Deduplication
Duplicate or overlapping room records are detected and reconciled.
Review workflows
High-value records and licensed outputs can receive manual review.
Delivery and refresh
Delivery formats
- +JSON
- +CSV
- +Parquet
- +REST API
Refresh options
- +Annual snapshot
- +Quarterly
- +Monthly
- +Real-time API