find_random_by

find_random_by_{filter} — same filter grammar as first_by, but orders randomly and limits to one row:

async fn find_random_by_team_id(&self, team_id: i32) -> sqlx::Result<Option<UserRow>>;
// SELECT * FROM "user" WHERE team_id = $1 ORDER BY RANDOM() LIMIT 1

The ORDER BY clause is backend-specific — RANDOM() on Postgres/SQLite, RAND() on MySQL — chosen at macro-expansion time from the active postgres/mysql/sqlite feature, so the method body itself is identical across backends.

Fetch mode is inferred from the return type the same way as first_by_*, including the same Vec<T> compile error (see first_by).

Not supported here: #[order_by(...)] (there's already a random order) — a compile error if present.