Unlike SQL, an aggregation over events produces a timeline indicating the value at each point in time over the events that have occurred up to that point in time. The query language provides composable, expression oriented syntax, including the ability to nest aggregations. Due to the focus on events and temporal queries, it provides capabilities to shift events forward (not backward, to avoid "leaking" information about the future into past values), ticks that create new times (allowing you to observe the current value of an aggregation every day, for instance) and joins that operate "at the current time".

Example query:

  # Purchases up to a point in time
  let purchases_now = count(purchase)
  # Shift that *forward* one day (so the value now is from 1 day ago)
  let purchases_yesterday = purchases_now | shift_by(days(1))
  # Compute the difference (purchases since yesterday)
  in purchases_now - purchases_yesterday