Tenant-safe by default
Scoped API keys establish clear boundaries before graph data is read or changed.
Graph persistence for application data
Drift
A compact, tenant-safe persistence service for things that already relate to one another: inventory, services, dependencies, documents, work, and more.
Focused by design
Connected data, kept ordinary
Drift handles graph-shaped persistence and stays out of the rest of the application.
Bounded, useful reads
Traverse relationships and retrieve declarative aggregates without unbounded queries.
Storage stays replaceable
SQLite is the first adapter, not the architecture. Core behavior depends on a stable repository port.
Why it exists
A persistence layer, not a graph platform
Most teams do not need a hosted graph product or an unbounded query engine. They need a small, stable layer that stores connected application data, preserves tenant boundaries, and leaves storage choices open as the system grows.
Guides & reference
Choose your path
Learn, operate, look up, or understand Drift without mixing those jobs together.
Tutorial
Build a first tenant-scoped graph from the ground up.
Open tutorialHow-to
Run, operate, release, and contribute to Drift.
Open how-toExplanation
Understand the boundaries behind adapters, retrieval, and tenancy.
Open explanationReference
Check the model, routes, scopes, and contract behavior.
Open referenceWhat it fits
The relationships already in your application
Services and dependencies
Model what runs where, what depends on it, and what a change can affect.
Inventory and assets
Keep equipment, products, ownership, and lifecycle data connected and queryable.
Documents and work
Represent related records without turning ordinary application data into a platform.