human review staffing calculator
Every agent workflow with a human checkpoint creates a queue. Divide the review hours by a working month and you get the naive answer, which misses the target every busy afternoon. This calculator uses Erlang C, the model contact centres staff with, to find the reviewers you need at the busiest hour to pick cases up on time, then adds shrinkage.
5 on shift at peak, 7.1 people in total
Erlang C queueing at the busiest hour. The gap between the naive workload figure and the headcount is the cost of answering on time, plus shrinkage. Cost assumes the peak headcount is held for every open hour, so it is an upper bound; staggered shifts bring it down. Occupancy above about 85% for long stretches burns people out.
questions
what is erlang c?
A queueing formula for work that arrives at random and is handled first come, first served by a pool of people. From the arrival rate, the handling time and the number of people, it gives the chance a case waits and how long. It is the standard way to staff contact centres and works the same way for review queues.
what is shrinkage?
The share of paid time people are not available to review: leave, sickness, training, meetings and breaks. 25–35% is common. Headcount is the reviewers needed on shift divided by one minus shrinkage.
why plan for the busiest hour?
Cases bunch up. If Monday mornings bring 1.6 times the average hour, staffing for the average means Monday's cases miss the target. Set the busiest-hour factor from your own arrival data.
what if the queue can wait a day?
Then set a long pick-up target, such as 480 minutes. The answer falls towards the plain workload figure, because a queue with a slow target can absorb peaks.
more free tools
- ai roi & token cost estimatorMonthly Claude API cost, cost per task and ROI for a workload
- eu ai act compliance auditorRisk tier, obligations and dates for one AI system
- agent workflow fit scorerIs this process a good fit for AI agents, and what shape?
- ai build vs buy calculatorThree-year cost: SaaS tool, custom agents, or keep it manual
- ai accuracy proof calculatorHow many test cases prove a vendor's accuracy claim?
- ai model routing calculatorWhat sending each request to the cheapest capable model saves
- ai vendor scorecardWeighted comparison of AI vendors, with must-haves and a swing check
- gdpr dpia screener for aiDoes this AI use of personal data need a DPIA?
Want the numbers checked against your own process? Book a 20-minute call.