Payment Observability & SLOs

From one normalized event to auth-rate, settlement-lag, and decline-mix SLOs

Payment Observability & SLOs From one normalized event to auth-rate, settlement-lag, and decline-mix SLOs 01 / Source 02 / Pipeline 03 / Aggregates 04 / Consume 05 / Act Payment Events · state transitions · 01 / Source · one ID Payment Events state transitions one ID Metrics Pipeline · roll-up job · 02 / Pipeline · sliced Metrics Pipeline roll-up job sliced Auth Rate · per slice · 03 / Aggregates · ratio Auth Rate per slice ratio Settlement Lag · aging buckets · 03 / Aggregates · p95 Settlement Lag aging buckets p95 Decline Mix · by reason · 03 / Aggregates · stacked Decline Mix by reason stacked Dashboards · trend + finance · 04 / Consume Dashboards trend + finance Alerting · SLO breach · 04 / Consume · paged Alerting SLO breach paged On-Call Action · routed owner · 05 / Act On-Call Action routed owner events append-only ratio settle time reason mix SLO metric SLO metric tech-error share threshold breach page Legend primary data policy / PII async batch data store

One Event, Many Metrics

  • • Every state transition emits one normalized event with a stable payment ID
  • • Auth rate, settlement lag, and decline mix are just aggregations over that stream

Alert on Symptoms

  • • Per-slice auth deviation, contractual settlement aging, and rising tech-error share page a human
  • • Each alert routes to the owner implied by its category, not a generic on-call bucket

Data-flow diagram • Built with Archify • Create yours ↗ • Hover to trace • R route • Click to focus • +/− zoom • M radar • T theme • E export