The integration of multi-laser additive manufacturing (AM) revolutionizes the printing process by facilitating concurrent printing, thereby considerably minimizing overall production time without sacrificing print quality. In multi-laser AM, decomposition methods and path planning strategies significantly influence the temperature distributions during the printing process. Thermal history is a critical factor, as it directly impacts the microstructures and thereby affects the mechanical properties of the printed component. For example, in regions such as intersection zones, inappropriate path planning can reduce the cooling rate of deposited material, resulting in prolonged elevated temperatures. This, in turn, can negatively affect tensile and yield strengths. Despite these known effects, a gap remains in understanding the combined impact of decomposition and path planning on thermal distributions in multi-head printing systems. Addressing this gap is essential for identifying decomposition and path planning configurations that produce uniform thermal profiles and ultimately enhanced part properties. To bridge this gap, this study introduces a novel framework leveraging physics-informed neural networks (PINNs) which is designed to be adaptable to various part shapes and geometries. The first step in the framework is efficiently decomposing a desired part using an equal-area constraint to ensure balanced segmentation. Following this decomposition, it proceeds to generate path for each of the subdivided sections for continuous paths solving it as a traveling salesman problem to enhance overall efficiency and precision. Laser power and scan speeds are then assigned for each section, with synchronization achieved by interpolating laser positions and their associated time steps. Eventually, a clustering model is developed based on statistical analyses of decomposed path plans, enabling the selection of desirable path-planning strategies. The selection criteria are based on the distance between the heads during the print. The clustering model's performance is confirmed by comparison with our novel path-aware PINN thermal model. The path-aware PINN is specifically designed to evaluate the thermal history for multi-laser print trajectories. This is achieved by minimizing the residuals of the governing partial differential equations and relevant boundary conditions. The proposed framework's flexibility in accommodating different geometries enables it to provide a desired solution space for decomposition and path planning in multi-laser printing processes. Furthermore, as the selection criteria for decomposition and path planning are integrated and aligned with temperature distributions, this framework presents a promising approach to enhancing the mechanical properties of dual-head additive manufacturing.