Automation and Abstraction in Many-Body Physics: Bridging Complexity and Computations

With an increasing demand for materials with target applications, it has become necessary to create automated workflows to search for optimal materials. While ground state properties which only require total-energy calculations can be streamlined efficiently, properties which require many-body interactions are harder to streamline due to complex workflow.

A typical calculation of an excited state property includes multiple steps where each step requires manual intervention. These manual interventions impede the search for optimal materials. In this work, we present a method of abstraction and the baseline infrastructure to streamline such many-body calculations for the search of optimal materials. We show this capability on a set of 200 2D materials by calculating their phonon limited carrier mobility. Finally, we present how this abstraction helps in performing computationally challenging non-equilibrium Green's function based calculations for carrier transport.