housing-search

Real-estate search and ranking system built for Datathon 2026 (ETH Zurich) and the RobinReal challenge. It turns natural-language housing queries into hard filters, soft preferences, geolocation signals, and ranked Swiss listings through a FastAPI backend, React frontend, MCP integration, semantic embeddings, and optional Bedrock/OpenAI LLM stages.

AI-assisted Manchester housing search PoC using preference capture, geospatial enrichment, soft ranking, tradeoff scenarios, and explainable property recommendations.

This repository contains the R code used to process the data, generate all results, and produce the figures presented in the document Who is Serious? A novel behavioral approach for identifying serious search on online real estate platforms.

Trova casa lungo la Linea 1 della metro di Torino: ricerca automatizzata su immobiliare.it con parcheggi live e indicatori di zona, mappa interattiva inclusa.