Bottleneck Detection
Bottleneck detection is the automatic identification of where delivery is being held up (a stalled review, a blocked ticket, a Work Unit that has stopped moving) so it can be addressed before it derails a deadline.
A bottleneck is any point where work accumulates faster than it clears. In software delivery, bottlenecks are usually invisible until they are expensive: a pull request waiting four days for review, a ticket blocked on an unanswered question, a Work Unit whose progress has flatlined while the deadline keeps approaching.
Bottleneck detection surfaces these conditions automatically by watching the rate and recency of activity across every Work Unit. Rather than waiting for a person to notice that something has gone quiet, LiveIndex flags Work Units whose cycle time is degrading, whose blockers have gone unresolved, or whose forecast has begun to slip, and explains why, with traceable links back to the specific PR or ticket responsible.
The value is leading rather than lagging signal. A missed deadline is a lagging indicator; a bottleneck forming three weeks out is a leading one. Catching the second prevents the first.
In LiveIndex, detected bottlenecks appear in the internal dashboard for the engineering team, and can optionally be summarized at an executive level on client reports so stakeholders understand risk without seeing internal detail.