Institutional dining facilities face significant operational challenges in balancing service quality with resource efficiency due to extreme temporal demand variability. This study addresses the staffing optimization problem for the Military Institute of Science and Technology (MIST) canteen using analytical queueing theory and discrete-event simulation. We employ the Erlang C (M/M/c) model to determine optimal server allocation across 34 fifteen-minute time slices spanning the operational day (07:30-15:45), with the objective of maintaining average waiting times below three minutes while minimizing excess capacity. Analysis of operational data revealed dramatic demand fluctuations, with arrival rates ranging from zero to 652 customers per hour and corresponding staffing requirements varying from one to ten servers—representing a 10:1 peak-to-trough ratio that precludes static workforce allocation strategies. The optimized dynamic staffing schedule achieves average waiting times of 0.35 minutes (analytical) and 0.40 minutes (simulated), representing 88% improvement over the service level agreement target. Validation through SimPy discrete-event simulation demonstrates strong agreement between analytical predictions and stochastic outcomes, with 94% of time periods validating within 10% tolerance and an overall correlation coefficient of 0.94. Sensitivity analysis reveals asymmetric system response to demand fluctuations: a 20% demand reduction yields proportional service improvements, while a 20% increase produces exponential degradation with three periods exceeding service targets despite maximum staffing. The lunch period (11:00-11:30) emerges as the critical capacity constraint, requiring full deployment of ten servers to manage peak arrival rates approaching system capacity limits. These findings provide actionable insights for institutional food service operators facing similar demand volatility challenges and demonstrate the efficacy of integrated analytical-simulation approaches for workforce optimization in time-sensitive service environments.