You can edit almost every page by Creating an account and confirming your email.

FalkorDB

From EverybodyWiki Bios & Wiki


FalkorDB
Developer(s)FalkorDB Ltd.
Initial release2023 (2023)
Written inC (originally; migrating to Rust)
Engine
    TypeGraph database
    LicenseServer Side Public License v1 (SSPLv1)
    Websitewww.falkordb.com

    Search FalkorDB on Amazon.

    FalkorDB is a graph database built as a module for Redis that represents graphs as sparse adjacency matrices and uses GraphBLAS linear-algebra operations to execute queries, rather than the pointer-chasing traversal used by most graph databases.[1] It implements the property graph model and the openCypher query language, and is marketed primarily as a low-latency knowledge-graph backend for large language model (LLM) applications, a technique commonly referred to as GraphRAG (graph-based retrieval-augmented generation).[2]

    History

    FalkorDB originated as a continuation of RedisGraph, a graph module for Redis first released by Redis Labs in November 2018.[3] After Redis (formerly Redis Labs) discontinued RedisGraph — ending new licenses in July 2023 and placing the project's GitHub repository into deprecation ahead of a January 2025 end-of-support date — a team of former Redis engineers forked the codebase and continued its development independently as FalkorDB.[3][4]

    The company behind the project, FalkorDB Ltd., was founded in 2023 by CEO Guy Korland, CTO Roi Lipman, and Chief Architect Avi Avni, all previously engineers at Redis.[5] In June 2024 the company announced a $3 million seed round led by Angular Ventures, with participation from K5 Global and a number of angel investors, alongside its participation in the Intel Ignite accelerator program.[5]

    Architecture

    Like its predecessor, FalkorDB stores a graph's adjacency information as one or more sparse matrices — a global adjacency matrix plus an additional matrix per relationship type — held in compressed sparse column (CSC) format, using the GraphBLAS library.[6] Multi-hop graph queries are executed as sparse matrix multiplications instead of sequential pointer lookups, which the developers say allows queries to be parallelized across CPU cores.[6][2]

    FalkorDB runs primarily as a Redis module reachable over the RESP protocol, with experimental support for the Bolt protocol used by other graph databases, and ships with a bundled browser-based query UI (FalkorDB Browser).[6] Beyond openCypher graph queries, it provides native full-text search, range indexing, and vector similarity search, allowing graph and vector operations to be combined in a single system.[7] A hosted, multi-tenant offering, FalkorDB Cloud, is also available, alongside self-hosted deployment via Docker or Kubernetes.[7]

    The original RedisGraph codebase was written in C; FalkorDB's developers have since undertaken a rewrite of core components in Rust, described as targeting memory safety and performance rather than changing the underlying sparse-matrix design.[6]

    Use cases

    FalkorDB is marketed for applications involving highly connected data, including knowledge graphs and agent memory for generative-AI and GraphRAG systems, fraud detection, cloud/cybersecurity analysis, and recommendation engines.[2][7]

    Licensing

    FalkorDB is distributed under the Server Side Public License v1 (SSPLv1).[8] The SSPL permits free internal and commercial use, but requires that anyone offering FalkorDB as part of a hosted service to third parties release the complete source code of that service under the same license; a separate commercial license is available for organizations that do not want this requirement.[8] The SSPL is not approved by the Open Source Initiative or recognized as a free-software license by the Free Software Foundation or Debian, so FalkorDB is generally classified as "source-available" rather than open source, a point of some public debate given the company's own "open source" marketing.[6]

    See also

    References

    1. "FalkorDB/FalkorDB: A super fast Graph Database uses GraphBLAS under the hood for its sparse adjacency matrix graph representation". GitHub. Retrieved 16 August 2026.
    2. 2.0 2.1 2.2 "FalkorDB Graph Database: Architecture, Benchmarks, License". Atlan. Retrieved 16 August 2026.
    3. 3.0 3.1 "RedisGraph EOL: FalkorDB Migration Guide". FalkorDB Blog. Retrieved 16 August 2026.
    4. "FalkorDB". Database of Databases. Retrieved 16 August 2026.
    5. 5.0 5.1 "FalkorDB, a Startup by Redis Veterans, Raises $3 Million to Enhance Language Models". openPR. 19 June 2024. Retrieved 16 August 2026.
    6. 6.0 6.1 6.2 6.3 6.4 "FalkorDB — Graph Database". GDB-Engines. Retrieved 16 August 2026.
    7. 7.0 7.1 7.2 "FalkorDB Docs | Graph Database for GraphRAG, Cypher & Knowledge Graphs". FalkorDB Docs. Retrieved 16 August 2026.
    8. 8.0 8.1 "FalkorDB License". FalkorDB Docs. Retrieved 16 August 2026.