Stellarators possess an immense amount of freedom in the choice of magnetic field. Modern stellarators are designed using numerical optimization techniques to find configurations with improved confinement properties consistent with engineering constraints. Navigating through the associated high-dimensional, non-convex design space is practically challenging. Our research focuses on the development of improved algorithms for efficient optimization techniques, such as through the deployment of adjoint methods. We contribute to the open-source SIMSOPT stellarator optimization and modeling software. We are applying these tools to the development of new experiments at the Columbia Plasma Physics Laboratory.
