Data Sources
Where the numbers in your report come from
Listing data
We fetch listings directly from the source portal — currently FincaRaíz and Metrocuadrado in Colombia. Photos, descriptions, prices, and amenities are read at the moment you paste the URL, so reports reflect the most recent snapshot.
Neighbourhood + walkability
OpenStreetMap (OSM) and Overpass for nearby cafés, transit, supermarkets, parks, gyms, pharmacies, and hospitals. Distances feed our Daily Life Score; the underlying data is updated continuously by the OSM community.
Safety + crime context
Where municipal data is available we use it directly — Medellín open crime data is the first source wired in. Elsewhere we rely on a combination of police-published statistics, expat community signal, and proximity-based heuristics. We always tell you which inputs were used.
Price fairness
Comparable rents are pulled from our own database of recently analysed listings in the same neighbourhood. The benchmark grows denser as more reports run, and we surface the comparables we used so you can sanity-check them.
Internet + connectivity
Address-level fibre availability where ISPs publish it, plus typical neighbourhood speeds inferred from public broadband datasets. We flag low-confidence estimates rather than guess.
AI analysis
Anthropic Claude reads the listing copy, photos, and the structured data above to produce the verdict, scam-risk signals, and the Spanish landlord message. The AI never invents prices or distances — those come from the data sources above.