IndiaReal estate2024 – ongoingHousing.com
Machine learning for fraud detection and property recommendations
Housing.com lifted its prospect-to-lead rate by 5.5% with personalized property recommendations built on machine learning
Problem
Housing.com's data sat in silos that did not match between teams. The company struggled with pricing accuracy, demand forecasting, personalization and fraud detection. Its cloud data warehouse was also costly.
Approach
Housing.com moved from its cloud data warehouse to a lakehouse on the Databricks Data + AI Platform. It trained and deployed models with MLflow, including a property recommendation engine and a model to detect fraudulent credit card transactions. More models and a fraud-detection chatbot were still in development.
Results
Takeaway
Housing.com tied a gain in lead conversion to personalization. The page gives no baseline for the figure.
Categories
- Industry
- Real estate
- Business function
- Marketing and salesFinance, risk and legal
- Innovation type
- Product and service innovation
- Adoption
- Off-the-shelf tool or platform
- Evidence
- Vendor case study
Sources
Source summaries
Databricks describes how Housing.com, an Indian real estate platform, moved to a lakehouse and built machine learning models for fraud detection and property recommendations. It reports a 5.5% rise in the prospect-to-lead rate from personalization and a 0.05% cut in fraudulent card transactions. It also gives platform figures, such as 50% lower total cost.
Full details are in the original sources.
