Product-domain modules
Identity, authorization, entitlements, notifications, media, submissions, and audit workflows share a consistent tenant model instead of behaving like unrelated projects.
Launch Rail combines source-owned domain services, shared engineering conventions, and a customer-cloud operating model. Your team starts from a coherent platform and keeps control of the code and deployment.
Platform capabilities
Each lifecycle, dependency, interface, and deployment target is published in the product registry, so technical teams can distinguish what ships now from what is in pilot or planned.
Identity, authorization, entitlements, notifications, media, submissions, and audit workflows share a consistent tenant model instead of behaving like unrelated projects.
Available services publish REST and Connect/gRPC interfaces from versioned contracts, with documentation and API references linked from the service registry.
Transactional boundaries, PostgreSQL persistence, event delivery, retention, and provider choices stay visible in the architecture your team owns.
Structured logs, OpenTelemetry conventions, health checks, migrations, and service configuration follow shared patterns across the ecosystem.
The AWS path is being polished with customer-run deployment artifacts. GCP remains visible as a planned target until it reaches the same release gate.
The Agent Gateway pilots selected MCP tools behind tenant scope, allowlists, authorization checks, budgets, approval rules, expiry, and audit events.
Interface model
REST and Connect/gRPC remain the primary application interfaces for supported services. MCP access is presented through the Agent Gateway pilot, where tool exposure and mutations can be governed instead of making each service a public agent endpoint.
REST for broad application compatibility
Connect/gRPC for typed service calls
CloudEvents and NATS patterns for workflows
MCP through tenant-scoped Agent Gateway policy