Sinyi Realty's AI Intelligent Recommendation System personalizes property matching

Overview

Sinyi Realty, Taiwan's leading real estate solutions provider, built an AI Intelligent Recommendation System (launched May 2022) on the Databricks Data + AI Platform and Delta Lake, using machine learning to power its Property Recommender Solutions and match properties to buyers based on consumer-centric behavioral and demographic attributes. The system delivered a 200% increase in website click-through rates, a ~30% increase in customer conversions, and a 20% increase in property closing rates.

The challenge

Sinyi Realty's roughly one petabyte of property, behavioral, demographic and session data sat in disparate, complex-to-scale legacy Microsoft SQL Server data warehouses without a single source of truth, leaving data teams unable to efficiently build pipelines or gain accurate insights for decisions.

The solution

Sinyi Realty modernized its data infrastructure on the Databricks Data + AI Platform and Delta Lake, unifying data into a common view to power its in-house Property Recommender Solutions and Property Matching System, which uses machine learning to deliver real-time, personalized property recommendations to customers, with Unity Catalog improving efficiency for analysts, scientists and engineers.

Reported business value

Since launching its AI Intelligent Recommendation System, Sinyi Realty saw a 200% increase in website click-through rates, a ~30% increase in customer conversions and a 20% increase in property closing rates, driven by a 7% increase in property impressions and 3% increase in click-through rate that together produced a 20% surge in the recommendation page's conversion rate.