Evolving Additive Manufacturing (AM) systems with multi-axis robotic arms and turntables are currently used to manufacture large metal-based parts. Based on the geometry, AM parts are realized either with planar, non-planar, or multi-planar material deposition strategies [1], as shown in Fig. 1. AM parts with overhangs manufactured using the planar strategy may require support structures during manufacturing and to reduce thermal-induced distortion. Non-planar strategy can be employed to minimize or eliminate support structures and simultaneously reduce thermal-induced distortion [2]. However, the non-planar strategy requires a continuous change in process parameters which may be less favorable for quality control. The multi-planar deposition is a compelling strategy for problems related to overhangs, support structure, and quality control. In this work, we present a mathematical method to optimize the multi-planar deposition strategy for metal AM parts to minimize thermal-induced distortion. The AM part is discretized using finite elements and each element is associated with a pseudo-time variable [2]. Through prescribed time intervals the AM part is segmented into sequential sub-parts using the pseudo-time variables. Each sub-part is associated with a local build direction which is also to be determined using numerical optimization. Based on the build directions, the sub-parts are further segmented into orthogonal planar layers with constant thickness. The sequential deposition of the constant thickness planar layers mimics the multi-planar material deposition in the AM process. We present a differentiable formulation for optimizing the multi-planar deposition strategy and solve the problem using gradient-based optimization. The inherent strain method is employed to model the total AM part distortion [2, 3]. The finite elements associated with a given layer are given a constant inherent strain value to model the distortion associated with the deposition of that given layer. The numerical examples of AM parts with optimized multi-planar deposition strategy will be presented to showcase the applicability of the proposed approach.