- Enter the licensed learner seats. Use the maximum number of learner accounts or seats available for the period being modeled.
- Set operating hours. Enter the number of hours in a typical day when learners can use the service.
- Enter average session length. Use the expected active time per learner session in minutes.
- Set parallel session capacity. Enter how many learner sessions the service can support at the same time.
- Choose expected utilization. Use a realistic percentage below 100% to allow for uneven demand and idle capacity.
- Review the bottleneck. The result uses the lower of the license limit and calculated session throughput.
AI Tutoring Learner Capacity Estimator
This estimator calculates how many learners a AI tutoring program can realistically serve during a day or operating period when both account limits and session throughput matter. It combines the licensed learner ceiling with available operating time, average session length, parallel session capacity, and an expected utilization rate.
The result is useful when planning enrollment, purchasing seats, or checking whether expected demand fits the delivery model. Because the estimate is based on operating assumptions rather than a vendor guarantee, it is best used for scenario planning: change session length, parallel capacity, or utilization to see which constraint becomes the practical bottleneck.
Capacity assumptions
learners
hours
min
sessions
%
Result
—
estimated learners served per day
Throughput capacity—
Licensed seat limit—
Binding constraint—
Throughput capacity = (Operating hours × 60 ÷ Session minutes) × Parallel sessions × Utilization rate
Estimated learner capacity = min(Licensed seats, Throughput capacity)
Estimated learner capacity = min(Licensed seats, Throughput capacity)
Utilization rate is entered as a percentage and converted to a decimal. The calculation assumes each counted session serves one learner and that capacity can be reused across sequential sessions during the operating day.
What the result means
The main result is the estimated number of learners that can be served in one operating day under the entered assumptions.
Actual platform limits, peak-time demand, scheduling rules, outages, and session overlap can reduce realized capacity.
Given: 1200 licensed seats, 16 operating hours, 25-minute sessions, 40 parallel sessions, and 75% utilization.
Calculation: (16 × 60 ÷ 25) × 40 × 0.75 = 1,152.0 learners/day of throughput. Capacity is the lower of 1,200 seats and that throughput.
Result: 1,152 learners per day.
The binding limit in this example is session throughput, so increasing the other input alone would not raise capacity.
Why can the result be lower than my licensed seat count?
Licenses set one ceiling, but operating time and parallel session throughput can create a lower practical ceiling. The calculator reports the smaller of those two limits.
What utilization rate should I enter?
Use the share of theoretical session capacity you realistically expect to fill. If demand is uneven or schedules leave gaps, a value below 100% is usually more useful for planning than the theoretical maximum.
Does one learner count more than once if they have multiple sessions?
The throughput model counts session opportunities, so it works best when each learner is expected to use one modeled session in the period. If learners use multiple sessions per day, divide or reinterpret capacity accordingly.
What happens if session length changes?
Shorter sessions increase the number of sequential session slots that fit into the operating day, while longer sessions reduce it. The effect only matters when session throughput is the binding constraint.
Is this the same as concurrent user capacity?
No. Concurrent capacity is the parallel session input; learner capacity also accounts for how many session cycles fit into the day and the utilization assumption.