Forecasting process may become cumbersome under erratic demand structures where no conventional methodology seems to provide plausible outcomes. The company investigated in this paper is subject to this condition and needs to forecast the workload in order to plan the indispensable process enhancement activities. Two parameters to be forecasted are the total production volume and the ratio of in-house production ratio with respect to the total production. The required forecasts for relatively regular structured in-house production to total production ratio were computed via static forecasting. The more important erratic structured total production volume forecasts were computed using FFT (Fast Fourier Transform) algorithm. Total production was further observed during the next year to check the reliability of the forecasts. The resulting respective MAPEs (Mean Absolute Percentage Errors) of approximately 11% (in-house ratio) and 14-22% (total production) were considered acceptable under the erratic demand structure.