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 the core Feldera control plane and the Kubernetes layer that runs them. You'll design and build the APIs and services that customers depend on to deploy, manage, and scale feldera data pipelines, and you'll lead the enterprise-grade capabilities - authentication, SSO, multi-tenancy, and cloud integration - that let large organizations adopt Feldera securely on their own infrastructure. What You’ll Do Control plane engineering: Own and evolve the control-plane services that orchestrate pipelines across diverse customer environments. Kubernetes platform: evolve the Kubernetes layer that runs Feldera pipelines reliably, including resource management and self-hosted deployment workflows. Enterprise readiness: Build and harden capabilities to support security, isolation, and access-control needs of large enterprises. Cloud & self-hosted deployment: Develop the capabilities that let customers run Feldera across environments, from single-node laptops to multi-node clusters, on their own infrastructure. API design: Design clean, stable, well-documented APIs for the control plane that internal teams and customers build workflows against. Reliability & operability: Make the control plane observable, debuggable, and resilient at scale. Troubleshooting: Debug complex distributed-systems issues across the control plane and Kubernetes layer. Minimum
