About the role
About Feldera Feldera is redefining how engineers compute over changing data. Powered by an award-winning breakthrough in database theory (DBSP), our platform incrementally maintains even the most complex SQL views as data changes, even when pipelines have hundreds of joins, aggregates, unions, and even recursion. It has allowed leading enterprises to have always-on real-time insights over both live and historical data, with 10x lower cost and 100x faster time-to-production. In this role, you’ll focus on our secure, scalable, production-grade self-hosted platform; one that customers deploy on their own infrastructure, from laptops to multi-node clusters.
The Role We're looking for a Rust backend engineer to own and evolve Feldera's connector ecosystem. You'll build and maintain connectors with systems like Delta Lake, Apache Iceberg, Kafka, and Postgres. This means working deep in the details of each system's storage formats, wire protocols, transactional semantics, and change-data-capture mechanisms, and turning that into ingest and output adapters that are fast, correct, and resilient to failure. What You’ll Do Connector engineering: Own and evolve Feldera's input and output connectors for data lakes (Delta Lake, Iceberg), streaming systems (Kafka and the broader Kafka ecosystem), and databases (Postgres and other OLTP/OLAP stores). Change data capture: Build robust CDC ingestion that turns upstream changes into correct, ordered, and consistent change streams into Feldera pipelines. Connector framework: Design and improve the shared abstractions, APIs, and tooling that make authoring, testing, and operating connectors fast and consistent. Formats & protocols: Work at the level of table formats (Parquet, Delta, Iceberg metadata), serialization formats (Avro, JSON, Protobuf), and wire protocols to integrate cleanly and efficiently with each system. Correctness & fault-tolerance: Ensure connectors are reliable and operate without data loss or duplication beyond stated guarantees. Performance: Push connector throughput and latency to keep pace with demanding production workloads. Troubleshooting: Debug complex distributed systems issues spanning Feldera and external systems in customer environments. Minimum
