The soft-bodied, midwater polychaete Tomopteris is an interesting swimmer. Not only do Tomopteris swim continuously throughout their life, they also perform two modes of locomotion simultaneously: metachronal paddling and bodily undulation. Tomopteris have two rows of flexible legs (parapodia) positioned on opposite sides of its body. Although each row performs a metachronal beating pattern, they paddle out of phase to one another. Both of these paddling behaviors occur in concert with its lateral bodily undulation. This undulation appears to further displace the parapodia, assisting the metachronal paddling process. We created a self-propelled, fluid-structure interaction model of a Tomopteris to explore how these two modes of locomotion synergize to generate effective swimming. In particular, we studied performance holistically over a 6D parameter space (leg length, leg number, paddling amplitude, undulation amplitude, body width, and fluid scale) with our sights on investigating higher-dimensional parameter spaces. In today's talk, I will describe how we approach studying Tomopteris swimming, through a blend of computational fluid dynamics and machine learning techniques.