Skip to main content
There is no customer cluster, PrivateLink endpoint or installer. You sign in at chat.xpander.ai and your agents run. The runtime controls are the same as in every deployment: see Security Architecture.

​
Architecture

Cloud architecture. Users and AI Engineers reach xpander Cloud over TLS. Inside, a control plane holds xpander Chat, identity with SSO, the agent registry, and the skills catalog; a data plane holds the execution tier, the agent-controller's skill relay (drawn as AI Gateway), PostgreSQL, and Redis with vectors, hosting threads, memory, and knowledge bases. Model calls go out to LLM providers and skill calls go out to the systems your skills reach. Captions state TLS at xpander's ingress, encryption at rest with cloud-provider managed keys, and that skill credentials are injected by the skill relay and never visible to the model.

xpander cloud. xpander operates the control plane (xpander Chat, identity, agent registry, skills catalog) and the data plane (the execution tier, the agent-controller's skill relay, PostgreSQL, Redis). LLM providers and the systems your skills reach sit outside and are reached from the data plane.

The deployment_type field in the List Agents API still reads serverless.

​
Identity and access

RBAC governs who may build, deploy and use agents. API keys are scoped and revocable. Audit logs record skill calls, data access and configuration changes (the record). Every control on Govern lives in xpander Chat.

​
Encryption and data handling

xpander.ai is SOC 2 Type II certified (Security, Availability, Confidentiality), with the report under NDA. ISO 27001, HIPAA and FedRAMP are in progress. See Certifications & Compliance.

​
Next steps

Need data residency in your own network, or zero egress to xpander? Use a self-hosted deployment.

Hybrid Architecture

Data plane in your VPC

Air-Gapped Architecture

Zero egress to xpander

Review the runtime controls

Permissions, credentials, safety

Request compliance documents

SOC 2, GDPR