Forge Runs Locally
In Sovereign Mode, Forge (an on-premise open-weight small language model) handles all AI inference locally. There is no outbound call to Apex or any external API.
- --All detection verdicts produced on your own GPU hardware
- --Forge Handover recalls raw logs from the on-prem ring buffer
- --No API keys, no external model endpoints, no internet dependency
- --G-Score and Detection Stream populated entirely from local inference
Zero Outbound Connectivity
Sovereign Mode is designed for environments where the security data network has no internet path. GhostAnalyst does not require cloud connectivity to operate.
- --HUD served locally on the Sidecar host
- --ClickHouse deployed on-premise for all telemetry storage
- --Redis and all pipeline services run in the same isolated cluster
- --Suitable for critical national infrastructure and classified environments
On-Premise Everything
Every component of the platform, from collection to AI to the admin HUD, can run inside your perimeter with no cloud dependency.
- --Single Docker Compose or K8s manifest for full stack deployment
- --Air-gap delivery via tarball or private registry mirror
- --All OCSF telemetry stored locally, no egress, no SaaS billing
- --Fully auditable: all code and models available for inspection
Reclaim your security data.
60-day full-platform trial. No credit card. No inbound firewall changes. Live in under an hour.