Waiting Time Analysis Agent – Find idle time between process steps, spot bottlenecks and explain their causes
Finds the time in which nothing happens: idle time between steps, bottlenecks and the reasons why work sits still.
What does the Waiting Time Analysis Agent do?
The Waiting Time Analysis Agent finds the time in which nothing happens in your processes. In many business processes, idle time makes up a large part of the cycle time: work waits for an approval, for a reply, for the next batch run. The agent measures these waiting times between activities, shows where they pile up and uses process data and company knowledge to explain why work sits there.
Questions the Waiting Time Analysis Agent answers
“Where does work sit longest in the purchase-to-pay process before it moves on?”
“How long do invoices wait for approval on average – and at which cost centers the longest?”
“Are there weekdays or times of day when work piles up?”
“What share of the cycle time is waiting time?”
How the Waiting Time Analysis Agent works
- 1
Measure waiting times
For every transition, the agent calculates the time between the end of one step and the start of the next.
- 2
Locate bottlenecks
It ranks transitions by total waiting time – case count times duration – and finds the places where the process loses the most time.
- 3
Recognize patterns
It checks whether waiting times relate to weekdays, month ends, teams or document types – typical signs of batch processing or capacity limits.
- 4
Name the causes
Using policies and known issues from the Context Hub, it explains why work waits at this point and cites the source.
KPIs and methods
- Waiting time per transition
- Mean and median idle time between two activities.
- Total waiting time
- Sum of waiting time across all cases – shows where a lever pays off.
- Waiting time share
- Share of waiting time in the total cycle time.
- Time patterns
- Waiting times by weekday, time of day or part of the month.
A practical example
In purchase-to-pay, the agent sees that invoices wait a long time for approval after entry – and not evenly: waiting time rises sharply in the second half of the month and mainly affects invoices without a purchase order. From the Context Hub it knows the approval policy that routes such invoices to cost center owners, and so names the actual cause: an approval level with hardly any capacity during month-end closing.
Matching end-to-end processes
What you get
A ranking of the transitions with the highest waiting times, time patterns as a chart and an explanation of why work sits there – with references to the data and documents the agent relied on.
What data the agent needs
A perspective with timestamps per activity. With start and end timestamps the agent cleanly separates waiting and processing time; with a single timestamp per event it measures the time between events.
Waiting Time Analysis Agent: frequently asked questions
What is the difference between waiting time and cycle time?
Cycle time is the total duration of a case. Waiting time is the part of it in which nobody works on the case – it sits in a queue, an inbox or waits for an approval. Anyone who wants to reduce cycle times usually starts with waiting times.
How does the agent recognize a bottleneck?
A bottleneck shows when work piles up before a step and the waiting time there is clearly higher than elsewhere. The agent weights waiting time by the number of affected cases and checks whether it rises with the amount of open work.
Does the agent need start and end timestamps per activity?
Not necessarily. With start and end timestamps it cleanly separates waiting and processing time. With a single timestamp per event it measures the time between events and points out that this includes processing time.
Part of Minerva 2.0
The Waiting Time Analysis Agent is an agent in Minerva, the AI assistant in Noreja. It uses the same data sources, the same choice of model – local or cloud – and the same access rights: it only reads what you are allowed to see and changes nothing in Noreja.
More agents from the library
Cycle Time Analysis Agent
Measures cycle times per case, transition and variant, shows the distribution rather than just the average and explains the outliers.
View agentWIP Analysis Agent
Counts open cases per process step, shows their age and detects where a backlog is building up.
View agentRoot Cause Analysis Agent
Answers the question of why: traces dependencies in the causal process graph and explains problems with evidence.
View agentSee what the Waiting Time Analysis Agent finds in your process data
In a demo we show you what answers the agent delivers on one of your processes – with numbers, reasoning and sources.
Book a demo