Tenants
The isolation boundary for users, secrets, executions, and routing in AEGIS.
Tenants
A tenant is the fundamental isolation boundary in AEGIS. Every user, secret, agent, workflow, execution, webhook route, and credential lives inside a tenant. Two tenants on the same cluster cannot see one another's data, share secrets, or accidentally pick up each other's webhooks — isolation is enforced at every layer of the platform, from the database all the way out to the storage gateway.
A single AEGIS cluster can host many tenants side by side. When a person signs up for the consumer product, a tenant is provisioned automatically. Larger organizations get a tenant that maps to their workspace, with multiple human users and service accounts attached to it. Almost every API call, gRPC request, and event in the system carries a tenant identifier, and the orchestrator refuses to act on anything where that identifier is missing or mismatched.
A small set of operations live above tenants — installing the cluster, registering a node, running disaster recovery — and these are reserved for cluster-level administrators. Everything else is tenant-scoped by default.
Key ideas
- Tenant — the isolation boundary that owns users, agents, workflows, secrets, and executions.
- Multi-tenant cluster — one AEGIS deployment can serve many independent tenants concurrently.
- Tenant-scoped routing — webhook source names and credentials are keyed per tenant; collisions across tenants are impossible.
- Cluster-level operations — node install, upgrades, and platform-wide admin sit outside any tenant.
Learn more
- Multi-Tenancy — how tenant isolation is configured and enforced across the platform.
- IAM — how users and service accounts authenticate into a tenant.