Music Tribe — One codebase. Twelve brands. 35TB migrated.
Music Tribe, a global music equipment conglomerate, operated 12+ brand websites as isolated codebases — each with its own deploy pipeline, CMS, and infrastructure. Updates required coordinating across teams, assets were duplicated across regions, and the legacy cloud setup couldn't scale to their growing product catalog spanning millions of records.
Engineered a unified Next.js monorepo to serve all 12+ brand websites from a single codebase with SSR, ISR, and CSR rendering optimized per brand. Built a Pimcore-based DAM, PIM, and MDM system to normalize millions of unstructured CSV/Excel product records into clean relational models. Architected and executed a 35TB cloud migration (5.6M+ files) using a distributed Python pipeline with parallel workers, streaming transfers, idempotency checkpointing, and fault tolerance — achieving ~60% deduplication and reducing ~8.85TB of redundant storage with zero downtime. Owned the entire GCP infrastructure: load balancers, instance groups, Kubernetes worker clusters, Elasticsearch, and CDN. Designed a cloud-agnostic architecture portable to AWS via config-driven deployments, and secured all asset delivery using UUID-based file paths to prevent enumerable CDN URLs.
All 12+ brands served from one codebase with independent deploy capabilities. 35TB migrated across clouds with zero downtime and full data integrity. ~8.85TB storage reduction through intelligent deduplication. Millions of product records normalized and searchable via OpenSearch. GCP infrastructure running in production with AWS portability built-in.
