Among the main challenges regarding the food industry, it could be noted that it is necessary to make food products with the required quality and affordable cost. Computing modeling methods solved such tasks by optimizing food recipes based on specific product criteria. However, the problem of optimizing school meals still needs to be addressed. Cottage cheese and culinary products are valuable sources of protein and calcium. They are particularly in demand for schoolchildren’s nutrition. This research aimed to develop molded culinary products with cottage cheese (cheese pancakes) based on the criteria convolution method using the Harrington-Mecher desirability function, i.e., a generalized desirability function in order to meet the physiological needs of school-aged children. The components used were cottage cheese, two additives from Jerusalem artichoke: (M1: «Jerusalem artichoke, pumpkin, pear» or M2: Jerusalem artichoke, quince, mountain ash), egg, and dry components (barley flakes or oat flour). Components were combined at specific ratios; the masses obtained were subject to structural-mechanical and sensory analysis. To determine the optimal equation of the dependence of the critical shear stress (yield stress) regarding the quantitative recipe composition, regression analysis was conducted, followed by Multi-criteria optimization based on Harrington’s desirability function. Two cheese pancake recipes were proposed, as they showed optimum quality characteristics of the final product: Recipe 1 (%): Cottage cheese 74.0, plant additives 14.0, oat flour 7.7, eggs 4.3; Recipe 2 (%): Cottage cheese 69.0, plant additives 17.0, flakes barley 10.0, eggs 4.0. The proposed approach was effective in optimizing various molded culinary products and represents a methodological basis for finding new innovative technological solutions in the food industry and public catering.