A well-known consumer brand in Los Angeles is rebuilding its core systems. That means a new cloud ERP going live region by region, a new merchandise planning platform, and product lifecycle management after that. You'll build the warehouse, pipelines and data checks behind its inventory and sales numbers. Once planning moves to automated allocation, a wrong number moves real product, so the controls you build are what catch it first.
What you'll do
- Build and run the BigQuery warehouse: ingestion, Dataform models from raw data to business marts, cost and performance tuning, and day-to-day support
- Build the warehouse side of the ERP move: the new feed, history carried across phased cutovers, and reconciliation through a parallel run
- Build and run the feeds the planning platform acts on, checked before they’re sent and recoverable after a failed run
- Move data quality to the source with data contracts, freshness and completeness checks, tested tolerances, and reconciliation of sales and inventory across stores, ERP, ecommerce and logistics partners
- Route every data issue to the owner who can fix it in time to act
- Build the semantic layer for plain-language and AI-assisted querying, and act as second reviewer on what reaches the certified layer
- Work AI‑first with Claude Code or Codex across planning, design, build and review
What you bring
- 5+ years in data engineering, building and running a production cloud warehouse (modeling, testing, lineage, cost and performance tuning)
- Expert SQL and strong Python
- Tested transformation code in Dataform, dbt or similar, with code review
- Hands‑on agentic engineering with Claude Code or Codex on real work. This is a must.
- Quality checks that run before data is promoted: freshness and completeness gates, reconciliation to source, alerts to named owners
Nice to have
- An ERP migration or cutover with history and parallel‑run reconciliation
- Feeds into systems that act on data automatically, like planning or replenishment
- BigQuery, Dataform and Google Cloud; Spark or Databricks; medallion architecture
- Terraform or other infrastructure as code
- Working alongside implementation partners and business data owners
The details
- On‑site in Los Angeles, at least 3 days a week
- Full‑time, direct hire
- Starts as soon as possible
- US work authorization required; no visa sponsorship