Senior Data Platform Engineer

Tubi - Canada•Toronto, ON

1d•CA$116,000 - CA$235,100•Hybrid

About The Position

We're hiring Senior and Staff Data Platform Engineers to join the Data Infrastructure teams in Toronto. Together these teams own the infrastructure that processes billions of events per day: Spark-on-Kubernetes, Flink and Kinesis pipelines, a multi-petabyte Delta Lake, a large-scale MemoryDB feature store, Databricks multi-environment operations, and the catalog and lifecycle systems that govern it. The team is small and senior. Each engineer owns major platform components: you design it, build it, and support it in production. This is a hybrid-role based out of our Toronto office. You must be willing to travel to our Toronto office two days/week.

Requirements

3+ years building and operating production data platform infrastructure at the cluster or platform level, across Spark, Flink, Kinesis, Kubernetes, or equivalent
Deep experience in at least one of: Spark-on-K8s cluster operations, Rust-based data or systems engineering, Kubernetes platform engineering and IaC, or data catalog and governance tooling
Production AWS experience or equivalent: EKS, S3, Kinesis, and multi-account IAM patterns (EKS Pod Identity, KIAM, or IRSA)
You've owned a critical platform component, you wrote the runbooks, tracked the cost, and were on-call for it
Strong in at least one of: Rust, JVM (Java or Scala), or Python for data platform work

Responsibilities

Spark-on-Kubernetes — EKS-based compute platform for Spark workloads: cluster configuration, Pod Identity IAM, job environment setup, Kustomize overlays, and shadow canary validation
Event ingestion — Rust services and Flink jobs processing billions of events per day over Kinesis; throughput, reliability, on-call response, and AI-assisted operational tooling to reduce toil
Platform infrastructure — Terraform modules for environment provisioning, cross-account AWS IAM, ARC runner infrastructure, and CI/CD for data platform changes
Feature store and ML compute — Flink-based real-time feature pipelines feeding a large-scale MemoryDB cluster; GPU capacity governance and Databricks multi-environment operations for ML training workloads
Workflow orchestration and CDC — Airflow-based DAG deployment, change data capture pipeline operations, and data quality monitoring