Memgraph

Memgraph is an open-source, in-memory graph database designed for real-time streaming and graph analytics workloads, developed by Memgraph Inc. (founded by Dominik Visek, Tomislav Vrbanec, and Marko Budiselic). Unlike disk-based graph databases like Neo4j, Memgraph stores all data in RAM, enabling sub-millisecond query latency for traversals, shortest path calculations, and graph algorithms on datasets of billions of nodes and edges. Memgraph is fully Cypher-compatible, meaning it uses the same Cypher query language as Neo4j, allowing users to migrate queries, tools, and expertise between the two databases with minimal changes. The database integrates natively with Kafka, Redpanda, and Pulsar message streams for real-time data ingestion and continuous graph updates, making it ideal for fraud detection, network topology analysis, recommendation engines, and real-time risk assessment use cases. Key features include: in-memory storage engine with ACID transactions via an optimistic concurrency control protocol, horizontal scaling via replication (one primary and multiple replica nodes), the MAGE (Memgraph Advanced Graph Extensions) library providing over 50 built-in graph algorithms including PageRank, Betweenness Centrality, Community Detection (Louvain, Label Propagation), Shortest Path (Dijkstra, Bellman-Ford, A-star), Cycle Detection, and machine learning algorithms, Python and C++ query module extensions for custom algorithm development, trigger support for executing actions on data changes, full-text search integration, and a visual interface via Memgraph Lab (an Electron-based desktop application for exploring and querying graph data, with Cypher query editor, graph visualization, and query performance analysis). As of 2026, Memgraph has over 1,400 GitHub stars.