Step 11: Gaps, Roadmap & Production Hardening

This document outlines architectural trade-offs knowingly accepted in the current codebase and provides an engineering roadmap ranked by priority (P0, P1, P2) with concrete implementation suggestions.

---

Priority Engineering Roadmap

Priority	Feature / Gap	Current State	Production Target & Fix
P0	Distributed Message Backbone	Database table queue (queue_state).	Replace relational queue state with Apache Kafka or AWS SQS FIFO for infinite horizontal queue scaling.
P0	Distributed Rate Limiting	Process-local token bucket (TokenBucketRateLimiter).	Migrate token bucket to Redis Cluster via Lua scripts (EVALSHA) to enforce shared rate limits across worker clusters.
P1	Active Circuit Breaker	Deterministic fallback sorting based on reported health.	Implement sliding-window circuit breakers (CLOSED, OPEN, HALF-OPEN) using Netflix Hystrix or Resilience4j algorithms.
P1	Quiet-Hours Postponement Queue	Suppressed messages terminate in SUPPRESSED status.	Add delayed queue release: hold suppressed notifications and automatically dispatch when local morning window opens (e.g. 08:01 AM).
P1	Real-Time Console Push	Operations console polls backend endpoints periodically.	Implement Server-Sent Events (SSE) or WebSockets for zero-latency live delivery feed updates.
P2	Cold Storage Archival	Delivery attempts remain in primary relational table.	Implement daily pg_partman partition rotation; archive records older than 90 days into ClickHouse or Amazon S3 Parquet.
P2	Mutual TLS (mTLS)	API key authentication in HTTP headers.	Enforce mTLS certificates for all inter-service communication within the service mesh.