Tenant Records

Nile Database rebuilds PostgreSQL around the tenant, so each customer's data sits in its own space. A Beam agent reads tenant records, watches for new tenants and status changes, and applies your rule for onboarding and housekeeping. When a tenant is added, the agent populates default rows, tags the record by plan, and notifies the owning team. Tenants showing odd growth, data that crosses expected limits, or signs of a misconfiguration are routed to an engineer, since anything touching one customer's isolation is a call a person should confirm.

Row-Level Updates

Nile Database stores each tenant's application rows in standard Postgres tables. A Beam agent reads and writes those rows where the connector and permissions allow, watches for records meeting a condition, and applies the approved rule for updating them. On an event like a status change, the agent updates the affected rows and notifies the owner of what moved. Bulk changes across many tenants, deletes, or updates that would touch sensitive fields are held and routed to an engineer, so a person approves anything large or irreversible before it reaches live data.

Tenant User Provisioning

Nile Database keeps user accounts scoped to their tenant. A Beam agent reads user records, watches sign-ups and access changes, and applies your rule for provisioning and cleanup. When a user joins a tenant, the agent sets default access, tags the account, and notifies the team; when a user goes inactive, it flags the account for cleanup. Elevated permissions, cross-tenant access, or anything that widens who can see a customer's data are routed to an engineer, since access across tenant lines is exactly where a human should decide.