Deployment Options
📅 Last reviewed: February 2026
WeSoar deploys wherever your data lives — multi-tenant SaaS, dedicated private cloud, on-premises, or fully air-gapped. Every model delivers the same 85+ AI capabilities.
Four Deployment Models, One Platform
Regulated industries face fundamentally different data sovereignty requirements. WeSoar is architected to run identically across all four deployment models, so you choose based on your security posture, not your feature needs.
Multi-Tenant SaaS
Fastest time-to-value. Hosted on WeSoar-managed infrastructure with SOC 2 Type II compliance. Data encrypted at rest (AES-256) and in transit (TLS 1.3). Go-live in 1–2 weeks.
Dedicated Private Cloud
Single-tenant on dedicated infrastructure within your preferred cloud region (AWS, Azure, GCP, G42, STC Cloud). Customer-managed encryption keys. Go-live in 2–3 weeks.
On-Premises
Runs entirely within your data center on Kubernetes. All data stays within your physical perimeter. SSO via SAML 2.0/LDAP. Used by central banks and government entities across the GCC. Go-live in 3–4 weeks.
Air-Gapped
Complete network isolation for defense, intelligence, and critical infrastructure. AI models run locally, updates via secure media. Full 85+ capabilities with zero external dependency. Go-live in 4–6 weeks.
Infrastructure Requirements — On-Premises
Container orchestration: Kubernetes 1.24+ (OpenShift 4.12+, Rancher, or vanilla K8s)
Compute — application: Min 4 nodes × 16 vCPU, 64 GB RAM
Compute — AI inference: 2 GPU nodes (NVIDIA A100 or equivalent); CPU-only available at reduced throughput
Storage: 500 GB SSD min. PostgreSQL + S3-compatible object storage
Identity: SAML 2.0, OAuth 2.0/OIDC, or LDAP/AD
Frequently Asked Questions
What deployment models does WeSoar support?
WeSoar supports four models: multi-tenant SaaS, dedicated private cloud, on-premises within your data center, and air-gapped with zero internet connectivity. All deliver the same 85+ AI capabilities.
Can we start with SaaS and migrate to on-premises later?
Yes. WeSoar architecture is deployment-agnostic. Organizations frequently start with SaaS for speed and migrate to on-premises as requirements evolve. Data migration tooling is included.
What infrastructure is required for on-premises?
Kubernetes 1.24+ (OpenShift, Rancher, or vanilla K8s), minimum 4 nodes with 16 vCPU and 64 GB RAM, plus GPU nodes for AI inference. Detailed sizing guides provided during implementation.
How does WeSoar handle updates in on-premises deployments?
Containerized update packages delivered through a secure channel, applied during maintenance windows with rollback capability. Air-gapped environments receive updates via secure media transfer.
