Senior Data Engineer

Responsibilities

• Own the path from raw transactional and event data to trustworthy, well-modeled datasets powering BetMGM's analytics, ML, and operational systems.
• Design, build, and operate batch, micro-batch, and streaming pipelines feeding Snowflake — Prefect-orchestrated flows on ECS Fargate, dbt for transformation, Snowpipe Streaming and Kafka for event ingestion.
• Own the full dbt lifecycle (sources → staging → intermediate → marts) with model contracts, freshness SLAs, automated tests, and version-controlled documentation.
• Stand up Snowflake objects (warehouses, RBAC, resource monitors, Dynamic Tables, Iceberg tables) through Terraform — no Click Ops in production.
• Build AWS-native infrastructure for data workloads — S3, ECS Fargate, Lambda, EMR Serverless, Glue Catalog, IAM, Secrets Manager, VPC endpoints — entirely in Terraform.
• Maintain CI/CD pipelines (Git Lab CI or Git Hub Actions) that gate every change with linting, dbt build, unit tests, contract checks, and AI-assisted code review.
• Tune warehouse sizing, clustering, and query patterns for cost and latency; instrument credit usage via ; right-size before scaling up.
• Design RBAC, masking policies, and row-access policies that satisfy a regulated operator without becoming an access bottleneck.
• Own freshness SLAs and data contracts for the gold layer; triage incidents end-to-end.
• Direct AI coding agents as a force multiplier — writing specs, decomposing work, reviewing AI-generated PRs, and owning the architectural decisions agents cannot make.
• Partner with analytics engineers, data scientists, and ML platform engineers on shared standards (naming, testing, observability, lineage, cost attribution).

• BS or MS in Computer Science, Statistics, Math, or other STEM field — or equivalent practical experience.
• 5+ years building production data pipelines on a modern stack (Python + SQL + dbt + cloud).
• Deep Snowflake — beyond SQL into administration: warehouse sizing, RBAC, resource monitors, Streams/Tasks, Dynamic Tables, secure data sharing, cost tuning via .
• Strong AWS — S3, ECS/Fargate, Lambda, IAM, Secrets Manager, VPC — plus production experience with at least one of EMR Serverless, Glue, or MWAA.
• Terraform for both cloud and Snowflake — you have owned IaC, not just touched it.
• Orchestration fluency — Prefect, Airflow, or Dagster — and an opinion about when each is the right tool.
• CI/CD ownership — you have built quality gates that block bad code, not just YAML pipelines that pass.
• Bias toward outcomes — you describe past work in terms of SLAs, incidents, and customers served, not tool checklists.
• Nice-to-Haves:
  Snowflake-native ML (Snowpark, Cortex AISQL, Snowflake Notebooks) for in-warehouse scoring or unstructured workloads.
• Iceberg / open-table-format experience for cross-engine interoperability.
• Streaming experience — Kafka, Snowpipe Streaming, or Kinesis — with stated latency budgets.
• Reverse-ETL exposure (Hightouch, Census, or custom) into operational marketing or product systems.
• A demonstrable track record of shipping more with AI in the loop than without — not 'I have used Cursor,' but 'this is how I design work for an agent to do.'