Power Digital - Data Engineer, AI & Analytics

• 3+ years in data or analytics engineering, including 1+ years owning a dbt project of meaningful size in production, not just contributing models to one. • Advanced proficiency in Python and SQL, with a focus on production-grade code for data pipelines and modeling. • Deep expertise in dbt, not just writing models. Incremental strategies and full-refresh tradeoffs, Jinja and macros, packages, generic and singular tests, snapshots, source freshness, exposures, and how to keep a large project's DAG and materializations under control. • Strong command of Snowflake and the surrounding cloud data stack to operate autonomously as a foundational data owner. • Experience modeling in a multi-tenant environment, with the judgment to know when a client request belongs in a client layer and when it belongs in the core. • Working knowledge of marketing and advertising datasets. You've handled UTMs, attribution windows, and the gap between platform-reported and warehouse-reported conversions. • Proven experience designing and managing end-to-end data lifecycles from ingestion to serving, with reliability that holds for both BI and AI applications. • Familiarity with cloud-native infrastructure (GCP) and infrastructure-as-code principles. • Real adoption of AI-agentic development workflows (Cursor, Claude Code, GitHub Copilot) for coding, debugging, and system architecture. • Demonstrated ability to architect AI-ready data models (feature stores, clean semantic layers) that support downstream initiatives. • Experience with Git and CI/CD best practices, including automated testing you trust. • Comfortable shipping iteratively and refining data products based on live feedback. • Agency, consultancy, or services experience. Working across many clients with different stacks transfers directly. • Measurement work: incrementality, media mix modeling, attribution. • Server-side tagging, or retail and marketplace data. • You might be a fit if… • You're a data engineer at a brand or retailer and want to work closer to the decisions your data drives. • You've built marketing pipelines at an agency or martech company and know how they break. • You're an analytics engineer who wants to own the full system rather than just the modeling layer. • You've rebuilt your own workflow around AI coding agents and want that to be the job. • Key Performance Indicators (KPIs) • AI-accelerated development. Reduce median development time from approved requirements to production deployment for new data pipeline and modeling requests by 20% within the first 6 months, using the team's established baseline for comparable requests. Document and productionize at least 1 reusable AI-agentic development pattern within the first 90 days and at least 2 within the first 12 months, with each pattern adopted in at least 2 production workflows or projects. • Data quality and reliability. Maintain ≥99% accuracy and completeness across fields designated as critical for client-facing and AI-facing data products, measured through automated data quality tests and reconciliations. Maintain a ≥95% successful scheduled pipeline execution rate for owned production pipelines, excluding documented upstream vendor/platform outages. • Client request throughput. Deliver ≥90% of assigned client-bespoke modeling requests within the agreed-upon turnaround time, measured quarterly. Where a bespoke request introduces logic applicable across clients, evaluate and document whether it belongs in the shared core or client-specific layer for 100% of material modeling changes. • Cross-functional enablement. Launch or materially migrate at least 2 major production data assets within the first 12 months that support agency, client, product, or AI consumers. Within 90 days of each launch, demonstrate adoption through at least 2 active downstream consumers, applications, or teams per asset and achieve either a 20% reduction in related recurring data-support tickets or another pre-defined adoption/efficiency target agreed upon before launch. • Most Important Things (MITs) • Build and ship end-to-end data systems that enable AI features. • Deliver production-ready datasets and pipelines that unblock AI, product, and client teams. • Reduce fragmentation by building unified, AI-ready data foundations