Requirements
7+ years of experience in data engineering and software developmentAbility to write high-quality code in Java/Scala, Python or equivalent languagesDeep, hands-on production experience with Apache Flink - DataStream API and Table API / Flink SQL (core requirement)Demonstrated experience with Flink state management: keyed state, state backends (e.g. RocksDB), large state sizes and state TTLHands-on experience with checkpointing, savepoints and fault tolerance - exactly-once vs at-least-once semantics, recovery and savepoint-based job upgradesStrong grasp of event-time processing: watermarking, windowing strategies, allowed lateness and late-data handlingExperience diagnosing and resolving backpressure - parallelism, operator chaining and network buffer tuningExperience operating Flink on self-managed infrastructure (Kubernetes or YARN) - application vs session mode, high availability and rolling upgradesPractical experience with stream processing (Kafka Streams or equivalent) and messaging systems for high-volume workloads, including exactly-once sinks and schema registry usagePractical experience with distributed query engines (e.g. Trino/Presto or similar)Practical experience with ETL / data integration tools (e.g. Datastage, Informatica, Apache NiFi or similar) and SQL-based transformation frameworks (e.g. dbt)Strong SQL skills and understanding of data modeling and data warehousing for analytical workloadsHands-on experience with real-time / low-latency analytical stores (columnar or OLAP engines, e.g. Apache Pinot/ClickHouse or similar)Practical experience with big-data platforms and distributions (e.g. Cloudera, Hadoop ecosystem or similar)Practical experience containerizing and operating data workloads (Docker; Kubernetes a plus) and workflow orchestration tools (e.g. Airflow)Familiarity with data lake table formats (e.g. Apache Iceberg), data governance / cataloging tools (e.g. DataHub) and lakehouse management systems (e.g. Apache Amoro)Familiarity using AI tools for development and debugging (Claude, Cursor, Codex)
Description
Build high-scale, low-latency streaming data pipelines deployed on infrastructure we run ourselves (on-prem), not managed cloud services. You will design and operate high-volume real-time data systems end to end, with Apache Flink as the core stream-processing engine.