Software Development Engineer, Compute Platform

What you'll do

• Develop and maintain data-driven features for a large-scale batch compute platform.
• Own features end-to-end from initial data analysis and model building to production serving.
• Implement changes across the full stack, including platform services, control planes, and node-level configurations.
• Design and run production validations using shadow modes, staged rollouts, and success metrics.
• Diagnose and fix system issues resulting from misbehaving model predictions or heuristics.
• Analyze cluster utilization and capacity to create demand forecasts and node pool configurations.
• Perform quantitative analysis to inform sizing, planning, and prioritization decisions for the team.
• Participate in an on-call rotation to provide same-day resolution for platform issues.

What we're looking for

• Strong programming skills in Go, Python, C++, or similar to design, debug, and test complex software systems.
• Working knowledge of distributed systems and operating system fundamentals.
• Experience instrumenting large-scale infrastructure and quantitatively analyzing CPU, memory, I/O, or network consumption.
• Ability to reason quantitatively about production data distributions, tails, and risk for running workloads.
• Track record of owning projects from ambiguous problem statements through production.
• Strong communication skills to make quantitative results legible and actionable for non-data specialists.
• BS in Computer Science/related fields with 3+ years experience, MS with 1+ years experience, or a PhD.
• Experience building or operating large-scale compute platforms, improving utilization, or using Kubernetes/Slurm (preferred).