Data Scientist, Dataset Assessment (Lead)

Job Title

Data Scientist, Dataset Assessment (Lead)

Location

50% Onsite Sunnyvale, CA & 50% Remote

Duration

4 to 6 months

About the Project

Reality Labs Supply Chain is standing up an NPI data foundation to support AI-enabled supply chain workstreams (product change management / PRQ qualification, capacity and Capex planning, in-market quality and incident management, and end-of-manufacturing planning). Before we build, we need an honest inventory and quality baseline of the datasets these workstreams depend on. This role owns that assessment.

What you'll do

• Build the authoritative inventory of datasets supporting NPI and supply chain processes — source systems, owners, refresh cadence, lineage, downstream consumers.
• Run structured stakeholder interviews with data owners, OPMs, manufacturing ops, quality, and engineering to capture how each dataset is actually produced and used, not just how it's documented.
• Score every dataset against the 7-dimension data quality framework [7 dimensions: ...], with evidence and a defensible rationale per score.
• Define and document the scoring rubric, interview guide, and intake template so the work is repeatable after the engagement ends.

Deliver a prioritized remediation backlog

which datasets are fit for AI/automation use today, which need work, and what that work is.

• Hand off a clear findings readout to Supply Chain leadership and to the incoming Data Engineer.

Minimum qualifications

• 5+ years in data science, data quality, or data governance roles.
• Demonstrated experience building a dataset inventory or running a data quality assessment at enterprise scale.
• Working knowledge of supply chain, manufacturing, or NPI processes — enough to interrogate a forecast, capacity, yield, or inventory dataset credibly.
• Strong SQL plus Python (pandas) for profiling and statistical quality checks.
• Excellent stakeholder interviewing and written synthesis skills; comfortable driving conversations with senior XFN partners.
• Able to deliver a scoped assessment in an 8-week window with minimal ramp.

Preferred qualifications

• Prior Meta experience; familiarity with Meta's data stack (Hive/Presto, Databricks, Scuba, Unidash) and with PLM/MES data (Teamcenter, factory MES).
• Experience with data catalog / metadata management tooling and formal DQ frameworks (DAMA DMBOK or similar).
• Hardware NPI or consumer electronics supply chain background.