The opportunity
Unity iAds is looking for a hands-on data infrastructure and platform engineer with proven leadership skills to lead and evolve our self-service data platform that powers petabyte-scale batch and streaming, near-real-time analytics, experimentation, and the feature platform. You’ll guide an already excellent team of senior engineers, own reliability and cost efficiency, and be the focal point for pivotal, company-wide data initiatives- feature platform, lakehouse, and streaming.
Own the lakehouse backbone: Mature Iceberg at high scale—partitioning, compaction, retention, metadata—and extend our IcebergManager in-house product to automate the lakehouse management in a self serve fashion.
Unify online/offline for features: Drive Flink adoption and patterns that keep features consistent and low-latency for experimentation and production.
Make self-serve real: Build golden paths, templates, and guardrails so product/analytics/DS engineers can move fast safely.
Run multi-tenant compute efficiently: EMR on EKS powered by Karpenter on Spot instances; right-size Trino/Spark/Druid for performance and cost.
Cross-cloud interoperability: BigQuery + BigLake/Iceberg interop where it makes sense (analytics, experimentation, partnership).
What you'll be doing
- Leading a senior Data Platform team: setting clear objectives, unblocking execution, and raising the engineering bar.
- Owning SLOs, on-call, incident response, and postmortems for core data services.
- Designing and operating EMR on EKS capacity profiles, autoscaling policies, and multi-tenant isolation.
- Tuning Trino (memory/spill, CBO, catalogs), Spark/Structured Streaming jobs, and Druid ingestion/compaction for sub-second analytics.
- Extending Flink patterns for the feature platform (state backends, checkpointing, watermarks, backfills).
- Driving FinOps work: CUR-based attribution, S3 Inventory-driven retention/compaction, Reservations/Savings Plans strategy, OpenCost visibility.
- Partnering with product engineering, analytics, and data science & ML engineers on roadmap, schema evolution, and data product SLAs.
- Leveling up observability (Prometheus/VictoriaMetrics/Grafana), data quality checks, and platform self-service tooling.
What we're looking for
- 2+ years leading engineers (team lead or manager) building/operating large-scale data platforms; 5+ years total in Data Infrastructure/DataOps roles.
- Proven ownership of cloud-native data platforms on AWS: S3, EMR (preferably EMR on EKS), IAM, Glue/Data Catalog, Athena.
- Production experience with Apache Iceberg (schema evolution, compaction, retention, metadata ops) and columnar formats (Parquet/Avro).
- Hands-on depth in at least two of: Trino/Presto, Apache Spark/Structured Streaming, Apache Druid, Apache Flink.
- Strong conceptual understanding of Kubernetes (EKS), including autoscaling, isolation, quotas, and observability
- Strong SQL skills and extensive experience with performance tuning, with solid proficiency in Python/Java.
- Solid understanding of Kafka concepts, hands-on experience is a plus
- Experience running on-call for data platforms and driving measurable SLO-based improvements.
You might also have
- Experience building feature platforms (feature definitions, materialization, serving, and online/offline consistency).
- Airflow (or similar) at scale; Argo experience is a plus.
- Familiarity with BigQuery (and ideally BigLake/Iceberg interop) and operational DBs like Aurora MySQL.
- Familiarity with BigQuery (and ideally BigLake/Iceberg interop) and operational DBs like Aurora MySQL.
- Experience with Clickhouse / Snowflake / Databricks / Starrocks.
- FinOps background (cost attribution/showback, Spot strategies).
- Data quality, lineage, and cataloging practices in large orgs.
- IaC (Terraform/CloudFormation)
Additional information
- Relocation support is not available for this position.
- Work visa/immigration sponsorship is not available for this position