 ProxyFL, proxy-based federated learning, is a novel approach to distributed machine learning that allows multiple institutions to share data while preserving individual institution's privacy. This technique uses proxies to enable efficient communication between institutions without requiring a centralized server. Additionally, it allows for greater flexibility in terms of model architectures, which can be tailored to the specific needs of each institution. Experimental results demonstrate that ProxyFL outperforms other federated learning techniques with significantly lower communication costs and stronger privacy guarantees. This article was authored by Shivam Kholra, John Fung Wen, Jesse C. Cresswell and others.