Director, Production Analytics

The Director, Production Analytics owns Production Analytics globally: every site, every product line, and every production function across Manufacturing Engineering, Manufacturing Operations, Supply Chain, Planning, and Warehousing / Material Management. The ambition is not better dashboards. It is that a division leader can tell whether they will hit rate before they miss it, and that when we argue about where to invest, the argument is settled by data rather than by whoever presents best.

• ◆Apply analytics to production’s biggest problems. Hold an evidence-backed view of what is worth the company's attention and in what order, rather than working a request list. Take the hardest two or three problems in Anduril's flagship programs and build analytics into them end to end, hands-on.
• ◆Make the various business units in Anduril decide earlier. Shift production reporting from retrospective to leading: risk surfaced two months before the gate, not status recited on the day of it. Adoption is the deliverable. A correct dashboard nobody opens has failed.
• ◆Turn what worked into the enterprise standard. Generalize proven capabilities into supported products, standardize the definitions the work proved out, and consolidate the duplicated reporting they replace. A standard that survives contact with a program running at rate is defensible. One written in a document is not.
• ◆Answer what it takes to scale. Own the analytical side of a ramp from low rate to high rate, and keep a current, evidenced view of the constraints that actually bind as the company grows.
• ◆Build the team and the foundation. Lead the consolidated analytics organization and the managers within it. Build on and in the production process/team rather than beside them. Hold the line that the various business units own master data quality while the analytics team builds the tooling that makes defects visible and cheap to fix.
• ◆Your work should produce numbers the business can trust. Where a metric is known to be wrong, it is a defect with an owner and a date, not a caveat on a slide. Fix the data before building on top of it.

• ✓10+ years in analytics, data, or quantitative operations, including 5+ years leading analytics or data teams, and experience managing managers or clear readiness to do so.
• ✓A rigorous, teachable approach to ambiguous problems. You can describe how you decide what to work on, how you structure an analysis, how you know when you're wrong, and how you install that method in a team.
• ✓Genuinely hands-on. Strong SQL and a modern data stack, able to get into a dataset personally, find the answer, and defend the method. This role does not succeed as pure management.
• ✓Demonstrated ability to identify the highest-value problem in a messy operational environment and solve it analytically.
• ✓A track record you can point at: specific operational outcomes your work caused, with the numbers, the program, and the skeptics named.
• ✓Experience shipping analytics as supported products with real users, and retiring what they replace.
• ✓Able to influence business unit and functional leadership peer-to-peer without authority, and to work credibly with a program team that starts out unconvinced.
• ✓Must be eligible to obtain and maintain a U.S. security clearance. U.S. person status required per ITAR/EAR.

• ✓Operational depth in manufacturing, supply chain, or planning, including fluency with routings, work orders, MRP exceptions, and cycle counts.
• ✓Analytical ownership of a production ramp from low rate to high rate.
• ✓Experience building an analytics function from a scattered starting point, consolidating people, standards, and tooling that previously sat under separate owners.
• ✓Databricks and Palantir Foundry, dimensional modeling and semantic or ontology layers.
• ✓High-rate or high-mix manufacturing, defense production, or another environment scaling volume fast.