The field of macromolecular design has advanced rapidly in recent years, driven by powerful deep learning tools such as AlphaFold and coupled with highthroughput experimental platforms. These tools have enabled unprecedented speed in iterating through in silico design and experimental validation. This thesis explores the development of tools for designing both RNA and proteins. In the first chapter, I present Nucleologic, a Monte Carlo tree search algorithm inspired by the Eterna community, for the automated design of RNA sensors capable of computing functions such as logic gates. In the second chapter, I introduce Sculptor, a generative design framework that integrates a protein generative model and rotamer interaction field to create de novo protein binders targeting user-defined epitopes. The third chapter describes two smaller RNA efforts: 3DRNA, a deep learning method for RNA inverse design on fixed backbone using voxelized local environments; and RNAGym, a benchmarking suite for evaluating RNA models across tasks in fitness, secondary structure, and tertiary structure prediction.