HippoRAG is a retrieval-augmented generation system inspired by human long-term memory. It enables LLMs to continuously integrate knowledge from external documents by combining knowledge graphs with personalized PageRank for improved associative retrieval. The framework improves multi-hop reasoning and sense-making in RAG pipelines while maintaining cost and latency efficiency. It competes with other graph-based RAG approaches like GraphRAG, RAPTOR, and LightRAG.