In classical drug discovery, researchers sift through a large database of known compounds, until they find one that has a desired effect on a target virus/protein. A new technique, Inverse Molecular Design flips the script: Researchers specify a desired behaviour/effect, and Generative AI builds the drug molecules. Generative models for drug design navigate a configurational space vastly larger than their counterparts in image generation, necessitating massive volumes of highly specific training data. LIMA is the first MD software build for generative AI. Unlike other MD engines, it plugs directly into existing technologies used to train AI models (PyTorch), making it possible to directly simulate drug discovery experiments. LIMA MD will enable companies to generate compounds not possible in a wet lab and put this information directly into Generative AI models. Why are we doing this? To help pave the way for a new generation of drugs, with fewer and well known sideeffects.