The NLRP3 inflammasome is a sensor in the immune system that detects danger signals and triggers inflammation. Its overactivation drives harmful inflammation seen in diseases such as gout, type 2 diabetes, Alzheimer’s disease, and rare autoinflammatory syndromes. There is currently no approved drug which targets NLRP3. We utilised REINVENT4, a deep generative reinforcement learning framework, to enable the fast generation of novel small-molecule structures targeting NLRP3. Molecules were filtered to obtain a pool of structurally diverse high-affinity molecules, covering different pharmacophoric fingerprints. From these, we identified two promising candidate ligands that balance docking affinity, CNS properties and physicochemical profile.