Under the feed is a recommendation engine with two jobs most marketplaces never attempt: choosing who should
judge each new vehicle, and choosing what each person sees next. The community keeps the final say; the
engine makes it fast and fair.
Matchmaking, not a poll
Like a dating app, the engine decides who evaluates what. It balances category fans with outsiders, puts
urgent vehicles first and keeps a share of random exploration, so no single group can capture a verdict.
Confident verdicts from few votes
A Bayesian decision model updates after every vote and stops as soon as the verdict is statistically
confident. Unclear cases go to curators, so no owner waits indefinitely.
Voters earn their weight
Each voter’s influence is learned from how often they agree with confident outcomes. Speed-voting,
answer streaks, brand-new accounts and coordinated devices are discounted automatically.
A feed that learns taste
Watch time, saves, shares, offers and verdicts build a decaying taste profile across category, body
style, make, decade, price band and country. Ranking blends it with freshness and quality, with
exploration built in.