Responsibilities

• Automate core processes by taking recurring workflows — reporting refreshes, data loads, spend QA, and campaign operations — from manual to hands-off, with guardrails and rollback paths that make unattended automation safe on live spend.
• Give the team its hours back by measurably reducing analyst toil, with hours saved tracked and reported as a first-class KPI of this role.
• Advance our media-mix model by hardening it into a versioned, validated product with out-of-sample validation as standard practice, refresh automation, prior and sensitivity checks, and calibration of response curves against experiment readouts.
• Advance testing and channel allocation by industrializing incrementality testing — geo-experiment tooling, always-on test pipelines, and clean readouts — so budget-allocation decisions across channels are gated on measured lift, not platform-reported numbers.
• Raise the dashboarding bar by shipping a standardized, exec-quality dashboard system with templates, brand styling, and automated refresh so every team deliverable is self-serve, current, and something leadership actually opens.
• Own the team's Claude Code setup end-to-end including project conventions, reusable skills, custom agents, MCP servers, and permission allowlists.
• Convert one-off analyst workflows into shared, documented, versioned skills with dry-run defaults and confirm gates on anything that writes to a live ad account.
• Run onboarding and office hours, pair with analysts to level up prompting, tool-building, and review habits, and be the go-to person for anyone stuck on an AI workflow.
• Evaluate and integrate new capability safely including new MCP connectors, scheduled agents, and workflow orchestration.
• Automate recurring pipelines including reporting refreshes, warehouse cache loaders, data-quality checks, and spend-delivery QA on live geo tests.

Requirements

• Minimum 3 years in a technical marketing or data role such as marketing engineering, GTM engineering, marketing ops with real code, analytics engineering, or growth engineering.
• Strong Python and SQL — production code others run, with error handling, logging, and rollback paths.
• Hands-on production experience with at least two major ad-platform APIs including Meta, Google Ads, TikTok, Reddit, or Impact.com, including OAuth and token lifecycle management.
• Experience running write automation against live spend including budget, targeting, and state changes via API, with concrete examples of a rollback needed, a partial failure recovered from, and how unattended writes are gated.
• Daily-driver experience with agentic AI coding tools such as Claude Code or equivalent — has built skills, agents, and MCP integrations, not just chatted with a model.
• Cloud fundamentals including AWS Secrets Manager, Lambda, S3, and IAM concepts, or equivalent.
• Authorized to work in the United States without sponsorship.

Preferred

• Snowflake, dbt, or similar warehouse tooling; GA4/BigQuery exports.
• Familiarity with marketing measurement including MMM, incrementality and geo testing, and attribution.
• MCP server development and workflow orchestrators such as n8n, Airflow, or Step Functions.
• MMP and affiliate plumbing including Branch, AppsFlyer, and Impact.com postbacks.
• E-commerce or DTC hardware and subscription business context.
• Data-visualization craft — readouts an executive will actually read with hierarchy, annotation, and restraint.
• The enablement gene — would rather make five people three times faster than be the hero, and can explain technical guardrails to non-engineers without condescension.