HelixDB is an OLTP graph database that combines native vector search and full-text search capabilities, built in Rust on object storage.
HelixDB addresses the fragmentation problem in AI application infrastructure by consolidating multiple specialized databases into a single platform. Rather than maintaining separate systems for application data, relational queries, vector embeddings, and graph relationships, the tool unifies these concerns under one data model. It operates primarily with a graph-plus-vector approach but also supports key-value, document, and relational data patterns. Queries are authored through language-specific DSLs in Rust, TypeScript, Go, or Python and sent directly to a running instance via HTTP, eliminating build or deployment steps.
The tool suits teams building AI agents and retrieval-augmented generation systems that need to manage both structured relationships and semantic similarity searches without operational overhead. It is particularly valuable for applications requiring federated access to company data for agent memory and knowledge graphs. The CLI includes a bootstrapper called helix chef that can scaffold a complete project, seed example data, and hand off to supported AI agents for rapid prototyping. Developers can choose between interactive setup or manual configuration depending on their workflow preferences.
Maintainers respond to issues and pull requests within hours. Work in the issue tracker centers on bug fixes, enhancements, and documentation.
