Senior Site Reliability Engineer

Job Description

Join the Data Management (DM) team at the Vera

C. Rubin Observatory, one of modern astronomy's defining missions. The Rubin Observatory is a new astronomy facility in Chile designed to create a 10-year time‑lapse map of the southern sky through the Legacy Survey of Space and Time (LSST). As part of this team, you'll design, operate, and sustain the systems that process Rubin's data in near real time. LSST will generate 15 TB of raw pixels per night with its 8-meter mirror and 3.2‑gigapixel camera, creating one of the most demanding petascale data challenges in science.

The Data Management System's Prompt Processing Framework identifies and distributes Alerts for every astrophysical object that moves, changes, or appears in the sky within minutes of observation. These alerts include potentially hazardous asteroids, supernovae, and entirely new classes of transient phenomena. Your work will directly enable astrophysical discoveries by keeping Rubin's alerts flowing.

You will join a distributed team of roughly 80 scientists and engineers building and operating Rubin's petascale data management systems. Our work spans large-scale image processing, distributed databases, and production services. Python is our lingua franca, and we develop our software openly on Git Hub under an open‑source license.

You will own the reliability and robustness of Rubin Observatory's Prompt Processing Framework, the system responsible for detecting and distributing near‑real‑time alerts for transient and moving objects in the night sky. The Prompt Processing Framework runs on Kubernetes, with event‑driven scaling using Kubernetes Event‑Driven Autoscaling (KEDA) integrated with Redis Streams. It interfaces with PostgreSQL databases and Kafka to ingest data and publish alerts to the global astronomy community.

• Ensure, through both architecture and practice, the reliable operation of the near‑real‑time data processing pipeline and timely delivery of alerts to downstream brokers.
• Design and develop software that reduces operational risk and improves system resilience, scalability, and usability, including addressing failure modes, error handling, and contention in shared resources.
• Improve system performance and resilience by applying architectural and systems‑level optimizations to increase throughput and reduce end‑to‑end latency.
• Operate Dev Ops‑oriented continuous deployment of services using modern distributed systems tooling and development practices (e.g., Kubernetes, Helm, ArgoCD, Kafka, Redis).
• Develop monitoring dashboards and alerts for the prompt processing service and work with teammates to design and implement a sustainable on‑call rotation that provides coverage during the start of observing hours in Chile (typically 2‑5 pm Pacific Time), with limited off‑hours responsibility.
• Define KPIs and metrics for observability and accountability of the pipeline.
• Participate in the collective engineering activities of the team, including performing code reviews, acting as a troubleshooting buddy, participating in design discussions, and writing documentation to effectively capture and communicate architectural and implementation choices.
• Collaborate with members of the Data Management team to identify opportunities to improve tools, workflows, and operational practices.
• Share responsibility with the broader team for the overall success of the Data Management system, beyond the Prompt Processing Framework.

The Prompt Processing Framework is built on a modern, cloud‑native foundation. It runs on Kubernetes, with deployments managed via Helm and ArgoCD, and uses event‑driven scaling through KEDA and Redis Streams. The system integrates with PostgreSQL and Kafka to ingest data and distribute alerts, with additional databases including Cassandra and InfluxDB. Our primary development language is Python, and our code is developed openly under an open‑source model.

• Bachelor's degree and eight years of relevant experience, or a combination of education and relevant experience designing and operating distributed systems at‑scale in…