ArcFlow
The spatial-temporal engine for the physical world
Query space. Query time. Query what happens next.
ArcFlow is a purpose-built graph database where space and time are first-class dimensions. It doesn't bolt spatial features onto a row-based engine. It was designed from the ground up for workloads where position, movement, and prediction matter more than text search or key-value lookups.
What changed
The physical world generates spatial data at a scale that traditional databases weren't built for. 22 players tracked at 60fps. Thousands of spatial events per match. Trajectory projections that need to resolve before the next frame arrives.
Traditional databases bolt spatial features onto row-based engines. The result: 12ms queries when you need sub-millisecond. That gap isn't a minor inconvenience. It's the difference between real-time and too late.
ArcFlow was built from scratch for this workload. Graph + Vector + Temporal + Spatial + Reactive, all in one engine. No extensions. No adapters. No compromises.
Runs everywhere
Same engine, different scale
Native, Docker, Astro, mobile, edge, browser.
Every ArcFlow instance is a fragment that operates independently and syncs when connected. Run it on a stadium server processing 22 players at 60fps. Run it on a laptop during development. Run it on a phone for field scouting. The queries are identical. The engine is identical. Only the scale changes.
ArcFlow is going open-source under BSL-1.1. The engine is free. OZ's advantage isn't the database; it's everything above it: the hardware, the AI models, the operational playbooks, the...