{"slug":"helping-more-people-find-their-next-dream-home","url":"https://findausecase.com/use-cases/helping-more-people-find-their-next-dream-home","title":"Helping more people find their next dream home","description":"Housing.com, a real estate platform, moved from a cloud data warehouse to the Databricks Data + AI Platform, using Delta Lake and MLflow to train and deploy machine learning models for personalized property recommendations and fraud detection; additional ML models for pricing prediction and demand forecasting are currently under development. The fraud detection ML model lowered fraudulent credit card transactions by 0.05%, the recommendation engine drove a 5.5% increase in prospect-to-lead conversion, and the company cut total cost of ownership by 50% while reducing ML and pipeline deployment time by 10-15%.","company":"Housing.com","industry":"Real Estate","aiCapabilities":["Fraud & Anomaly Detection","Recommendation & Personalization"],"technology":["Delta Lake","Unity Catalog","MLflow","Delta Sharing","Tableau"],"deployment":"Public Cloud","problemStatement":"Data silos across teams threatened accuracy, and Housing.com struggled with pricing accuracy, demand forecasting, personalization and fraud detection as data volumes scaled; moving away from a cloud data warehouse, the cost of computing was proving far more expensive than storage.","solutionApproach":"Housing.com moved from its cloud data warehouse to the Databricks Data + AI Platform (paired with AWS), using Delta Lake as the foundational storage layer, MLflow to train and deploy ML models, Unity Catalog for governance, and Tableau for BI/reporting, to power a property recommendation engine and fraud detection; ML models for pricing prediction and demand forecasting are still under development.","businessValue":"Housing.com cut total cost of ownership by 50%, reduced pipeline/ML deployment time by 10-15% (saving one week of manual ML deployment work), increased team productivity/collaboration speed by 20%, lowered fraudulent credit card transactions by 0.05%, and increased the prospect-to-lead conversion rate by 5.5% via the recommendation engine.","evidence":{"band":"high"},"sourceUrl":"https://www.databricks.com/customers/housing-com","dates":{"publishedAt":"2026-10-05T05:48:26.921Z","publishedAtSource":"pipeline","updatedAt":"2026-10-05T05:48:26.921Z"},"license":"Open for reading and citing with a link to https://findausecase.com/use-cases/helping-more-people-find-their-next-dream-home. Bulk republication requires permission."}