Company brain
Connect documents, conversations, and decisions around customers, people, and projects so agents can follow context across tools.
HYDRADB / AI
Connected context for your agents.
HydraDB is a graph database built on object storage for applications that need rich, connected context. Connect facts, people, projects, decisions, and memories so your agents can follow relationships across sources.
This is the text-first version of the HydraDB website. Product context, applications, and integration resources are available here without interactive tabs or animations.
Two ways to build
- Open-source graph database. Self-host the engine using S3-compatible storage, openCypher, and Neo4j drivers. Follow the repository documentation for deployment and supported query features.
- HydraDB Cloud. Use the hosted context API to ingest knowledge and memories, build context graphs, and retrieve context for agents. Start with the agent integration guide.
What you can build
Multiplayer agent apps
Connect users, agents, tasks, decisions, and evidence around shared work so collaborators can continue with the latest context.
Ontologies
Model domain entities and relationships in a shared graph your agents can query alongside the facts and decisions attached to them.
Agent memory
Build a memory harness around users, tasks, and events so agents can retrieve relevant history across sessions.
Knowledge graphs
Build connected data foundations for ML, fraud detection, and supply chain analysis. Use relationship paths for features, investigations, and dependency tracing.
Start building
- Choose the self-hosted graph engine or the hosted context API.
- Read the corresponding repository instructions or agent integration guide.
- For the hosted API, create a database, wait for readiness, ingest content, wait for indexing, then query it.
- Use the OpenAPI specification for current request and response fields.
How it fits your stack
Use graph retrieval when your application needs to follow relationships across sources. Vector search helps find similar content; a graph helps follow the people, projects, decisions, and dependencies connected to it.
HydraDB can serve as the graph database behind a GraphRAG pipeline. Keep your models and extraction pipeline, then query entities and relationships for connected context.
Postgres with a graph extension can be a practical starting point. Separating graph workloads from your application database lets you scale graph queries without competing for its memory and CPU.
Resources
Agent integration guide ↗
Integration instructions for AI coding agents.
Documentation index ↗
Discover the available documentation pages.
OpenAPI specification ↗
Machine-readable schema for the hosted v2 API.
Quickstart ↗
Set up the hosted API and make your first query.
Open-source graph database ↗
Source code, deployment instructions, and Cypher compatibility.
Benchmarks ↗
Application evaluations, methodology, and results.
Deployment and pricing
Self-host the open-source graph database or use HydraDB Cloud. Dedicated deployments and enterprise support are available. See the pricing section for current plans, usage rates, and deployment options.