OpenAI· Workload· San Francisco
Site Reliability Engineer, Infrastructure - Analytics Platform
Comp$230K – $385K
Classified Tasks (36)
Automate 0%Augment 75%Human-Only 25%
Augment (27)
AI assists, human decides
Design, build, and operate critical infrastructure for research workloads
technical
Own production-critical infrastructure end-to-end
operational
Ship pragmatic solutions quickly
technical
Iterate on solutions using production feedback
operational
Own infrastructure lifecycle management across provisioning, upgrades, scaling, and decommissioning using an IaC-first approach
operational
Operate and scale ClickHouse clusters
technical
Implement sharding strategies for ClickHouse clusters
technical
Configure and manage replication for ClickHouse clusters
technical
Conduct capacity planning for ClickHouse deployments
analytical
Perform performance tuning for ClickHouse query and storage layers
technical
Execute maintenance tasks for ClickHouse clusters
operational
Operate Kafka as the ingestion backbone
technical
Improve Kafka throughput
technical
Reduce Kafka consumer and producer lag
technical
Implement backpressure handling strategies for Kafka pipelines
technical
Design and implement failure recovery for Kafka systems
technical
Maintain reliable integrations with Snowflake
technical
Improve end-to-end latency and reliability for data-heavy serving and query workloads
technical
Define and maintain SLIs and SLOs for services
analytical
Build and maintain monitoring dashboards
technical
Create and enforce alerting policies
operational
Author and maintain actionable runbooks
operational
Own backup and restore processes and disaster recovery strategy
operational
Plan and execute safe rollouts across dev, stage, and prod environments
operational
Implement canary deployment and rollback strategies
technical
Improve CI/CD pipelines and developer experience for faster, safer, and more predictable releases
technical
Strengthen security posture across infrastructure and delivery systems by implementing least privilege, secrets management, patching, and supply-chain controls
technical
Human-Only (9)
Requires human judgment
Translate ambiguous operational problems into actionable plans
analytical
Define and raise operational standards across teams
leadership
Perform hands-on operations and debugging in production systems
technical
Define and implement incident response standards
leadership
Establish and improve on-call practices
operational
Drive improvements in postmortem quality
leadership
Conduct regular disaster recovery and recovery drills
operational
Partner day-to-day with software engineers to embed reliability into design, implementation, and release processes
communication
Set operational readiness and runbook quality standards and drive adoption across teams
leadership
Job description
Site Reliability Engineer, Infrastructure - Analytics Platform | OpenAI Careers ## Site Reliability Engineer, Infrastructure - Analytics Platform Workload - San Francisco Apply now(opens in a new window) ## **About the Team** The Scaling team designs, builds, and operates critical infrastructure that enables research at OpenAI. Our mission is simple: accelerate the progress of research towards AGI. We do this by building core systems that researchers rely on - ranging from low-level infrastructure components to research-facing custom applications. These systems must scale with the increasing complexity and size of our workloads, while remaining reliable and easy to use. ## **About the Role** We're looking for an experienced Site Reliability Engineer to own production-critical infrastructure end to end. This role is centered on data-heavy, low-latency workloads, with emphasis on operating large-scale ClickHouse clusters, high-throughput Kafka pipelines, and reliable integrations with Snowflake. You'll turn ambiguous operational problems into clear plans, ship pragmatic solutions quickly, and improve them through production feedback and iteration. We are specifically looking for someone who can independently define and raise operational standards across teams while remaining deeply hands-on in production systems. ## In this role, you will * Own infrastructure lifecycle management across provisioning, upgrades, scaling, and decommissioning (IaC-first). * Operate and scale ClickHouse clusters, including sharding, replication, capacity planning, performance tuning, and maintenance. * Operate Kafka as the ingestion backbone, improving throughput, lag, backpressure handling, and failure recovery. * Improve end-to-end latency and reliability for data-heavy serving and query workloads. * Build and maintain strong monitoring and alerting: SLIs/SLOs, dashboards, alert policies, and actionable runbooks. * Define, implement, and continuously improve incident response standards, on-call practices, and postmortem quality. * Own backup/restore and disaster recovery strategy, including regular recovery drills. * Plan and execute safe rollouts across multiple environments (dev/stage/prod), including canary and rollback strategies. * Partner day to day with software engineers, embedding reliability into design, implementation, and release processes. * Set the quality bar for operational readiness and runbook standards, and drive adoption across teams. * Improve CI/CD pipelines and DevEx for faster, safer, and more predictable releases. * Strengthen security posture across infrastructure and delivery systems (least privilege, secrets management, patching, supply-chain controls). ## You might thrive in this role if you have * A track record of owning production infrastructure for data-heavy, low-latency systems end to end. * Strong hands-on experience operating ClickHouse, Kafka, and adjacent large-scale data systems. * Practical experience with Snowflake workflows and cross-system data architecture. * The ability to independently define operational standards (runbooks, incident process, rollout safety) and make them stick. * Strong operational experience with Kubernetes, Terraform, and cloud infrastructure. * Excellent communication and collaboration skills; you work effectively across engineering and research teams. * High personal rigor and organization in high-pressure production environments. * A deeply hands-on mindset: willing to debug incidents, tune systems, and implement fixes directly. **About OpenAI** OpenAI is an AI research and deployment company dedicated to ensuring that general-purpose artificial intelligence benefits all of humanity. We push the boundaries of the capabilities of AI systems and seek to safely deploy them to the world through our products. AI is an extremely powerful tool that must be created with safety and human needs at its core, and to achieve