DEPLOYMENT

Deployment Options

VIntercept runs entirely on your infrastructure — choose the right model.

Single-Node Deployment

All components run on a single GPU server — ideal for initial evaluations and smaller environments.

  • 1 server with NVIDIA GPU (A100, H100, or equivalent)
  • Suitable for up to ~5,000 monitored endpoints
  • Simplest deployment topology
  • Full platform functionality on a single node

Clustered Deployment

Components distributed across dedicated nodes for high availability and horizontal scaling.

  • 3+ nodes with GPU infrastructure
  • Horizontally scalable to 100,000+ endpoints
  • High availability with automatic failover
  • Dedicated nodes for ingestion, inference, and orchestration

Air-Gapped Deployment

Fully air-gapped with zero external dependencies for classified or regulated environments.

  • Zero external network dependencies
  • Offline model updates via secure transfer
  • FIPS 140-2 validated cryptography
  • Suitable for classified and SCIF environments
NEXT STEPS

Let Us Help You Choose

The right model depends on your size, compliance, and operational constraints — we determine it during the POC.