The article discusses NP-hard optimization problems of irregular cutting and problems of packing objects of arbitrary geometry. When placing objects, their representation in the form of orthogonal polyhedra is used, obtained after applying voxelization algorithms to given models of objects of complex shape. A greedy heuristic for the placement of two- and three-dimensional orthogonal polyhedra is proposed, which ensures the selecting of the best orientation variant for each object to be placed when searching for the densest placement scheme. The analysis of the effectiveness of the application of this greedy placement heuristic on the problems of flat irregular cutting and packing of three-dimensional objects of non-regular shape is carried out. The application of the proposed greedy heuristic provides fast obtaining if high quality solutions to the problems of packing a large number of objects represented in the form of orthogonal polyhedra. It is shown that the best solutions are obtained as a result of the joint application of the genetic algorithm and the proposed heuristic.