Redfin uses machine learning to automatically send personalized home listing recommendations to buyers and sellers at key moments during their search — functioning as an automated marketing system that works on behalf of Redfin agents. This is one of Redfin's earliest shipped AI products and drives roughly 25% of all Redfin website traffic.

The recommendation engine analyzes buyer behavior — which homes users click on, view, favorite, or tour — to predict which new listings they are most likely to engage with. Emails are triggered automatically by the model at moments it identifies as high-intent points in the buyer's journey. The system outperforms users' own saved search filters at predicting relevant listings, according to Redfin's CTO. Agents do not manually compose or send these emails; the system operates automatically within their client workflows.