Lead response to the most serious data incidents, drive blameless postmortems, and make sure the fixes stop the whole class of failure, not just the one instance.
Take part in the team's PagerDuty on-call rotation.
Lead performance, scale and cost.
Own capacity planning and load testing for seasonal peaks across the pipelines.
Lead deep performance work on Kafka partitioning, Flink parallelism, state and checkpointing, and Druid/StarRocks ingestion and query serving.
Set the architecture for Experiments, Remote Configs and Segmentation: assignment, sticky enrollment, targeting, exposure logging and results pipelines.
Partner with IXP and Data Science on statistical correctness and metric definitions.
Lead across org boundaries.
Represent Optimize in technical decisions with Data Platform, Streaming, Engine and ESP on shared infrastructure, such as the Powerhouse pipelines, QaaS-managed Kafka and the creator compute cluster.
Build alignment and drive multi-team initiatives to completion, influencing without authority.
Multiply the team.
Mentor senior engineers and grow the next technical leaders on the team.
Set engineering practices for design quality, operational excellence and AI-assisted development.
Partner with engineering, product and design leadership on planning and prioritization.
Must-have qualifications (or equivalent)
10+ years of professional software engineering, with 5+ years designing, building and operating large-scale data systems in production.
A track record of setting technical direction at the principal/staff-plus level: defining architecture across multiple teams or systems and seeing it through to measurable results.
Streaming: deep, production expertise with Apache Kafka and Apache Flink (or equivalent, e.g. Kafka Streams, Spark Structured Streaming, Beam/Dataflow), including event time, watermarks, windowing, state and exactly-once/at-least-once trade-offs.
Batch and orchestration: Apache Spark / Hive (or equivalent warehouse/lakehouse processing) and Apache Airflow (or equivalent DAG orchestration such as Dagster or Prefect), including DAG design, SLAs, dependencies and backfills at scale.
Analytical serving stores: Apache Druid and/or StarRocks (or equivalent real-time OLAP, e.g. ClickHouse, Pinot), including ingestion specs, rollups, segment/partition design and query tuning.
Shared data platforms: experience shaping a company-scale shared platform similar to Roblox's Powerhouse pipelines, i.e. shared ETL frameworks, standardized tables and multi-team clusters.
Backend services: expert in at least one of C#/.NET, Java/Scala, or Go, plus Python for pipelines and tooling. Experience designing and operating gRPC/REST services at scale.
Operational leadership: has defined SLOs, monitoring and alerting (Grafana/Prometheus-style), and led the response to and postmortems of major production incidents.
Influence and communication: clear, persuasive writing (strategy docs, tech specs, postmortems) and a record of building alignment across teams and with senior leadership.
Sponsorship data shows employers' past filings in public US government records (DOL, USCIS, E-Verify) and is not a promise that any job will be sponsored. Immigration rules change. Verify your situation with your DSO or an immigration attorney.