Machine Learning System Design Interview #1 - The Auto-Bidder Paradox

Why symmetric loss turns your model into a toxic-asset machine - and how quantile-based losses fix it.

You’re in a final-round interview for a Quant/ML role at the top Prop-Tech firm Zillow. The interviewer puts a chart on the table.

The Scenario: “We trained a Transformer-based regression model on 10 million home sales. It achieves an RMSE of 1.5%, significantly beating our human appraisers. We want to auto-bid on $500M of inventory next month.”

The Const…