ADR 0007 — Enterprise Zuora-class scope; Chargebee-primary upstream

Date: 2026-07-30
Status: Accepted
Deciders: RevRec Next design session (grill lock)

Context

Product aim is a Zuora Revenue–class outcome set, not a thin demo. Chargebee is the native upstream (~90% of designed source paths). Existing RevRec/revanu customers must be migratable. Implementation sequencing is out of scope for this ADR; capability scope is in scope.

Decision

1. North star: Zuora-class ASC 606 subledger + close + reports + audit (+ AR/AC sibling). Clone outcomes, not RevPro staging/RC-template architecture.
2. Upstream: Chargebee first-class adapter; extendable SourceAdapter interface; legacy file/Hotglue-shaped ingest = migration adapter emitting the same domain events.
3. Scope taxonomy (documentation): - Maturity labels: Foundation / Complete / Enterprise-deep / Out (see doc 16; aligns with prior MVP/Next/Later/Out). - Lenses: Zuora-class must-have · Chargebee 90% · Legacy migration must-have.
4. Correctness: ASC 606 + audit + close + journal integrity. Shadow compares use accounting equality. Smoke library is a policy/use-case quarry and validation aid for engine replacement — not sheet parity.
5. Enterprise NFRs are requirements: multi-tenant isolation, fat-tenant parallelism by contract, replay, period integrity, audit retention.

Consequences

• Full design (docs 13–17) describes the complete system, not a deliberately crippled model.
• Delivery can still slice vertically; the model and architecture must not omit multi-PO/SSP, mods, usage hooks, or close — they are part of the optimal design even if some policies land later in code.