ChrisCruz.ai Quarterly Planning

Q4 Planning Memo

Q4 Planning Playbook for Transaction History & CDP Migration

A unified narrative for Wil2 and BLR2 leaders to execute on data reconciliation, THV4 delivery, and CDP decommissioning while protecting performance, observability, and consumer trust.

September 20, 2025 • 8 minute read • By Christopher Manuel Cruz-Guzman
Executive Summary Key Insights Detailed Breakdown Implementation Guide Reflection Questions Atomic Notes

Executive Summary

Wil2 and BLR2 enter Q4 with a dual mandate: mature Transaction History Version 4 into a production-grade experience while exiting the CDP data center through AWS-native pipelines. The north star is simple—achieve 99%+ reconciliation, land THV4 with consumer-ready resilience, and harden observability before scaling traffic.

  • Data quality is the lead domino: reconciliation sits at 92%, driving a quarter-long push with SOR, ODS, and FCDP partners to close gaps.
  • THV4 represents a 100-point engineering investment spanning bug fixes, QA, pagination tuning, Cassandra resiliency, FMEA, and runbook creation.
  • BLR2 shoulders a 300-point quarter anchored in CDP decommissioning, AWS migration, and merchant tagging backfills—all under a critical timeline.
  • Cross-team friction points—API gateway decisions, alert fatigue, photon upgrades, and consumer onboarding—demand explicit owners and fast decisions.

Key Insights

Critical Decisions

“We need to make a quick decision… we’ve been sitting on this for a very long time.” — Kiran on the API gateway direction. Envoy remains aspirational; Q4 requires a call to unblock GLB creation and production readiness.

  • Decide between Envoy and existing gateway before sprint 2.
  • Align alerting scope so production signals stay actionable.
  • Stand up photon upgrade tracker for every API touchpoint.

Performance Expectations

THV4 must meet or beat V3 latency at heavy load. Targets call for 3K–5K TPS sustained throughput while tuning memory and CPU. BlazeMeter limits may cap perf env tests at ~3K TPS; the team will stress 2x expected load.

  • Document perf ceilings alongside mitigations.
  • Pair FMEA scenarios with telemetry validation.
  • Capture runbooks for incident hand-off prior to go-live.

Consumer Momentum

Wil2 is juggling V2 ➜ V3 migrations, digital channel onboarding, and aggregator education. An AI-powered onboarding agent is planned to parse documentation, videos, and FAQs—reducing support toil and accelerating adoption.

AI Onboarding Agent Swagger & API Store Alignment Cross-Product Consistency

Detailed Breakdown

Wil2 — Transaction History API

“It’s not about showing the numbers… identify each of the issues, fix the differences, work with the CDP team.” — Kiran on reconciliation discipline.

  • Data Reconciliation (Priority #1): Current metrics hover at 92%. The end-of-quarter target is 99%+, requiring sustained analysis of SOR vs. ODS discrepancies, validation of SOR fixes, and tighter loops with ingestion teams.
  • THV4 Development & Stabilization: Estimated at ~100 story points, covering pagination QA, defect burn-down, Cassandra retry resiliency, GLB creation, true CI/CD, breached-account routing strategies, and cross-product consistency validation.
  • Performance & FMEA Tracks: Performance testing breaks out as its own epic to tune latency and resource usage, while FMEA scenarios stress profile table dependencies and downstream integrations to confirm graceful degradation.
  • Consumer Integration Testing: Digital channels and aggregators are queued for ETD V3 testing; support stories remain open to triage defects and educate teams transitioning from V2.
  • Production Readiness: Observability, monitoring, alert tuning, Deadpool integration, runbooks, and swagger/API store parity are bundled together to reduce post-launch friction.
  • Operational Friction: Alert fatigue continues—“We’re getting hundreds of alerts… becoming numb to them,” Chris flagged. Photon upgrades need a centralized tracker, and the Envoy vs. existing gateway decision blocks routing design.

BLR2 — CDP Exit & AWS Migration

“CDP DC exit is a driving factor. AWS migration is a solution we came up with to do the CDP DC exit.” — Kiran

  • Migration Imperative: Decommission CDP and land workloads on AWS by year-end. Success promises faster processing and aligns with enterprise data strategy.
  • Data Quality Lift: Recon sits at 92%; leadership insists it is “unacceptable” to show externally. The goal mirrors Wil2 at 99%, with iterative back-and-forth validation alongside FCDP counterparts.
  • Major Workstreams: Merchant tagging historical loads, dynamic orchestration, consumer integration testing, account breach validation, CDP exit migrations, and SOD ingestion redesign (100 pts) all compete for focus.
  • Capacity & Resourcing: Q4 plan totals ~300 story points. A new AWS-experienced engineer joins October 1 to absorb migration load.
  • Risk Landscape: Tight timelines, monitoring readiness, data inconsistencies, and cross-team coordination top the list. Incremental migration and proactive comms anchor the mitigation plan.

Implementation Guide

4-Phase Q4 Execution Framework

  1. Stabilize the Core (Weeks 1-3): Lock the API gateway decision, formalize photon tracker, stand up reconciliation war room, and confirm perf environment limits.
  2. Build & Test (Weeks 2-6): Burn down THV4 backlog, execute BlazeMeter runs to 2x TPS, and complete FMEA scripts targeting profile table, downstream service, and queue failures.
  3. Harden & Document (Weeks 5-8): Finalize runbooks, calibrate alert thresholds, publish onboarding videos, and deploy the AI documentation agent prototype.
  4. Migrate & Monitor (Weeks 7-12): Sequence AWS cutovers, verify merchant tagging loads, track recon dashboards, and run cross-team go/no-go reviews.

Example Play — Recon Issue Response

1) Identify variance in dashboard, 2) assign SOR/ODS pairing owner, 3) validate upstream fix with FCDP, 4) rerun reconciliation batch, 5) archive resolution in tracker, 6) push update to AI onboarding corpus for consumer transparency.

This loop maintains the “identify & fix” discipline Kiran demanded.

Cross-Links for Deeper Context

Reflection Questions

  1. What’s blocking a final API gateway decision, and who owns clearing it this sprint?
  2. How quickly can we convert the 92% reconciliation metric into a daily 99% trend, and what joint rituals with SOR/CDP partners are missing?
  3. Where does alert volume still mask true incidents, and what telemetry do we turn off, tune, or automate through the AI onboarding agent?
  4. How will we confirm cross-product consistency (TH vs. Cross Product APIs) before production traffic doubles?
  5. Which AWS migration milestones need executive escalation now to protect the year-end CDP exit?

Atomic Notes

  • CDP exit portfolio includes Spark job migrations (175 pts), SOD ingestion redesign (100 pts), Cascade 2.5 uplift (50 pts), MMS updates (25 pts), DDA primary (205 pts), dynamic orchestration (30 pts), merchant tagging historical load (10 pts), ETD V3 consumer testing (10 pts), and Account Breach validation (30 pts).
  • Photon upgrades progressing toward v3.4.5; need confluence tracker to log version per API and ensure deployments accompany feature work.
  • Wil2 backlog sizing signals THV4 development (50 pts), performance testing (25 pts), FMEA (25 pts), production readiness (25 pts), and consumer integration (20 pts).
  • Alert fatigue surfaced via Splunk emails and build notifications; product stakeholders should be unsubscribed from BAU noise while engineering tunes signal-to-noise ratios.
  • Account breach flow may route to SOR for initial go-live (<0.1 TPS) to avoid complexity during THV4 launch.
  • New AWS hire onboarded Oct 1 with prior migration experience—deploy toward high-burn CDP exit user stories early.