Migration & ModernizationProduction

2021-2022

Korea Digital Platform Migration

Cutover project that moved orders, identity linkage, and data flows to the global standard

Project Overview

Migration project that transitioned Korea order, data, and messaging flows to the global standard while linking legacy user accounts to new global platform accounts through a shared user identifier so historical orders remained accessible after cutover. The same stream also stabilized legacy order history and post-purchase operations after go-live.

Challenge

The challenge was not only moving five years of order data on schedule, but also preserving account linkage, product consistency, clickstream continuity, and post-cutover legacy order operations at the same time.

Solution

Account-mapping-based order migration model

Designed the migration around a shared user identifier so order history remained valid under the new global account structure.

Designed the shared user identifier-based account linkage model
Implemented batch jobs for order transformation and linkage
Reduced migration processing time through performance testing
Built validation and operational plans aligned to cutover timing

Post-cutover legacy order and post-purchase integration

Organized the follow-up integration so historical order lookup, CS access, returns, and delivery status sync continued after cutover.

Connected legacy order lookup for nike.com
Handled about 20K additional order linkage tasks after cutover
Integrated CS lookup systems
Designed and operated returns and delivery-status synchronization

Tech Stack

Java

Implemented migration and validation logic

Spring Batch

Large-scale batch orchestration for order linkage and migration

Identity Mapping

Legacy-to-global account mapping model

Data Validation

Consistency checks and reconciliation

Cutover Planning

Cutover planning and runbook design

Prodigy

Global product information service integration

Clickstream

Global collection normalization

Bulk Email Ops

High-volume email risk management

Key Results

Completed cutover on 2022-10-04
Built migration structure for about 20M historical orders
Reduced migration processing time to about 3.5 days
Stabilized legacy order history and post-purchase operations after cutover

Learnings

Large migrations often hinge more on account-linking design than raw data movement
Cutover quality depends on performance, validation logic, and operational planning together
Product consistency and communication risks must be handled as part of the migration itself
Accurate structural framing increases credibility in public-facing case studies
Louis Kim - Software Engineer