Neutron transport is a central problem in nuclear physics and reactor engineering and is governed by the stationary Boltzmann transport equation, which expresses particle conservation in space, direction, and energy. The numerical solution of this equation using deterministic discrete ordinates formulations is computationally demanding and may exhibit convergence difficulties, particularly for refined angular discretizations and complex geometries. Classical schemes widely used in computational neutronics, such as Diamond Difference, Extended Diamond, Constant Nodal, and spectral nodal methods, provide high physical fidelity. However, when combined with angular sweeping and iterative procedures, these approaches may lead to significant computational costs in large-scale applications. In this context, this work presents an ongoing investigation into the feasibility of a hybrid approach that combines deterministic nodal methods with data-assisted techniques, with particular emphasis on angular order reduction and the potential acceleration of selected stages of the neutron transport solution. The proposed approach considers reduced-order models trained using synthetic data generated by well-established deterministic solvers formulated from the Boltzmann equation within the SN framework and spectral nodal methods. Techniques such as neural networks, autoencoders, and sparse system identification are explored as auxiliary tools to complement angular sweeping and iterative schemes, while preserving the physical structure of the governing equation. At this stage, the study focuses on conceptual analysis, identification of computational bottlenecks, and the definition of integration strategies between classical deterministic solvers and data-assisted models. No conclusive numerical results are reported at this stage, and the contribution is intended to assess the feasibility, limitations, and potential scope of hybrid nodal and data-assisted strategies for neutron transport problems.