Software Engineer, Observability

ABOUT THE TEAM

The Software Engineering team owns keeping our datacenters running at the scale and reliability required for frontier AI training and inference by building software systems and platforms to automatically predict, monitor, and rapidly respond to changing operational conditions and incidents. Our clusters are large, heterogeneous, and always on - the team makes them observable and keeps them running: we build the infrastructure that lets engineers and technicians expedite and automate their day to day, speeding node remediation and increasing overall GPU uptime. The team partners closely with datacenter operations, AI research, and facility infrastructure teams to deliver high-impact tools that turn raw operational data into clear insight and action, ensuring true site-wide reliability and automated operation.

RESPONSIBILITIES:

• Design and implement scalable, fault-tolerant data ingestion pipelines that reliably consume high-volume telemetry from compute nodes, GPUs and accelerators, schedulers, storage, network fabric, and facility systems (power, cooling, energy storage) in real time.
• Develop frontend visualizations, interactive dashboards, and UI components that present processed data in clear, real-time views for operators, capacity planners, and engineering stakeholders.
• Design, build, and operate multi-service production systems (UI, APIs, data pipelines, auth, and on-call) to enable reliability, uptime monitoring, predictive alerting, upgrade/change/risk management, and site-wide impact assessments.
• Build and maintain the software stacks that make site operations scalable, auditable, and fast - including repair trackers, vendor RMA workflows, operational infrastructure, and their integrations with ticketing, inventory/rack systems, telemetry stores, and vendor portals.
• Drive automation that reduces manual effort in cluster monitoring and improves anomaly detection accuracy.
• Keep the tools themselves reliable: uptime, data integrity, reconciliation, access control, and safe deploys.
• Work day to day with your users, in person, to rapidly incorporate feedback and process improvements that improve GPU uptime.
• Deep-dive into operational problems, find efficient solutions, and apply first-principles thinking.

 

BASIC QUALIFICATIONS: 

• Bachelor's degree in computer science, engineering, math, or a STEM discipline and 3+ years of professional software development experience; OR 5+ years of professional software development experience in lieu of a degree.
• Programming experience in Python, Rust, C++, Java, or similar languages.
• Experience working with Postgres, MySQL, MongoDB, DynamoDB, or similar database management systems.
• Proven ownership of correctness-sensitive systems - state machines, workflows, or operational data where inaccurate state has real-world impact.
• Experience collaborating with cross-functional teams.

 

PREFERRED SKILLS AND EXPERIENCE:

• Experience building or operating large-scale observability, telemetry, or real-time analytics systems.
• Experience with time-series or analytical databases (Prometheus/VictoriaMetrics, ClickHouse, InfluxDB) and streaming or high-volume telemetry pipelines.
• Experience building observability dashboards and alerting with Grafana, New Relic, Splunk, or similar.
• Frontend experience with React, Angular, or a comparable JavaScript framework, building intuitive UIs for critical operations.
• Experience designing and shipping multi-service systems for mission-critical applications - RESTful APIs, data stores, and at least one of: UI, pipelines, or auth.
• Experience operating services on cloud or on-prem infrastructure with Docker/Kubernetes in production and automated CI/CD deploys (ArgoCD, GitHub Actions, etc.).
• On-call discipline and a track record of owning internal platforms end-to-end with production rigor in rapidly changing environments.
• Experience with HPC or supercomputing environments (schedulers such as Slurm, GPU/accelerator telemetry, high-speed fabrics, or parallel filesystems).
• Experience integrating with industrial/OT data sources (SCADA platforms such as Ignition, tag historians, Modbus/OPC-UA) or power/cooling infrastructure telemetry.
• MS in Computer Science or a related field.

ADDITIONAL REQUIREMENTS:

• Ability to work extended hours or weekends as needed for mission-critical deadlines.
• Ability to work on an on-call basis on a weeklong rotational schedule.