Graph Engineering Has a New Name, But the Concept Is Decades Old

A 25-year enterprise infrastructure veteran argues that graph engineering — currently hyped through GraphRAG, code graphs, and entity graphs — is fundamentally a rebranding of entity-relationship thinking that infrastructure professionals have practiced for decades. At its core, graph engineering structures knowledge as nodes and typed relationships rather than flat documents, enabling systems to traverse connections and answer complex dependency queries. The author draws direct parallels between modern graph concepts and traditional infrastructure work, such as dependency maps, CMDB systems, and blast-radius impact analysis, noting these are essentially graph traversal problems. While acknowledging that specialized graph database engineering involves distinct technical skills, the author's central claim is that the underlying mental model is not new. The piece aims to demystify vendor-driven framing that makes practitioners feel behind on a concept many have long applied under different names.