One product · multiple boundaries
Forge AI goes
where Forge goes.
Forge Accelerant uses the same Web, API, PostgreSQL evidence model, tenant controls, and Forge Reasoner contract across managed SaaS, customer public cloud, private cloud, on-premises, Kubernetes, and OpenShift. The installation model changes where services run—not what Forge can do.
AI boundary: when Forge AI is enabled, the Forge API connects to a private Forge Reasoner endpoint. External model fallback is disabled by default. Customer evidence does not need to leave the selected deployment boundary.
Deployment matrix
What changes — and what does not
| Offering | Web/API | PostgreSQL | Forge Reasoner | Identity | Evidence boundary |
|---|