Enhancing Accuracy and Performance of Aligned Spin EOB Models with NRPy+ and BOB

As gravitational wave interferometers become more sensitive and detection rates increase, it is essential to develop more accurate and efficient compact binary merger waveform models. To this end, we present novel strategies to enhance the development, performance, and accuracy of spin-aligned waveform models based on the Effective-One-Body (EOB) formulation. We start by implementing our version of the state-of-the-art inspiral model, SEOBNRv5HM, using an intuitive Python infrastructure documented in Jupyter notebooks. To enhance performance, we leverage our Python package, NRPy+, to generate highly optimized C code. The optimized inspiral is smoothly attached to our in-house highly accurate Backwards-One-Body (BOB) merger-ringdown model, which is a first principles model based on the properties of the final merged remnant. Our approach combines meticulous documentation of the EOB physics with code optimization, resulting in a robust and efficient implementation. We conduct comprehensive performance and accuracy comparisons between the traditional SEOBNRv5HM model and our newly proposed approximant. We also outline future optimization strategies for gravitational wave approximants.