This document describes a proposed proximity-aware and interest-clustered peer-to-peer (P2P) file sharing system (PAIS) that forms physically close nodes into clusters and further groups nodes with common interests into subclusters. It aims to improve file searching efficiency by creating replicas of frequently requested files within subclusters. The system analyzes user interests and file sharing behaviors to construct the network topology and uses an intelligent file replication algorithm. The experimental results show this approach improves file searching performance compared to existing P2P systems.