Skip to content

Home

AkasicDB

“A Unified Graph-Vector-Relational DBMS for Enterprise AI”

Welcome to the official AkasicDB documentation. AkasicDB brings high-performance graph querying, vector search, and full-text search directly into PostgreSQL.

Key Benefits

  • PostgreSQL Integration Load the akasicdb extension to run graph and vector workloads through SQL.

  • Graph Generation from SQL Define graph structures and create graphs from relational data using SQL.

  • Graph-Vector-Relational Integrated Query Perform graph queries, relational queries, and vector search together in a single integrated query.

  • Multiple Vector Indexing Algorithms Choose workload-specific indexes such as HNSW, Vamana, IVF, Flat, and DiskANN.

  • Full-Text Search Search multilingual text, including Korean, with USING texton indexes and BM25 ranking, without a separate search infrastructure.

  • Administration Tune graph and vector execution with GUC parameters, and inspect graph and vector index state.

  • Client Packages Build applications with the AkasicDB Python library and LangChain integration.

  • Client Examples Use SQL, Python, and LangChain examples for client-side integration.