Own the design of the control plane that places, reconciles, and operates VMs across the Compute fleet
Define the database partitioning and sharding strategy to scale control plane state management 5x
Develop capacity models for placement and inventory state
Own the placement reconciliation loop, including desired-state modeling, idempotent reconcilers, drift detection, and reconciliation-lag SLOs
Generalize bulk and asynchronous operations into one model with standard operation resources, batching, concurrency caps, and automatic halt on failure thresholds
Set consistency guarantees per data path
Enforce idempotency and fencing on controller actions
Specify saga-style workflows
Requirements:
12 years of experience in building distributed systems, along with experience owning control planes or large-scale orchestration or equivalent
Deep database experience with SQL, including partitioning, sharding, replication, and online schema change
Experience building asynchronous systems, including durable task queues, event-driven workflows, and retry and backpressure design
Hands-on experience building and operating containerized applications, including Kubernetes
Competence in Python or Go
Experience with the other language or ColdFusion is good, but not required
Deep understanding of consistency, including leases, fencing, idempotency, and optimistic concurrency
Hands-on experience with reconciliation or controller patterns, such as Kubernetes-style controllers or equivalent
Benefits:
Annual bonus or incentives
Equity awards
Employee Stock Purchase Plan (ESPP)
Healthcare
401K savings plan
Company holidays
Vacation (in the form of PTO)
Sick time
Parental leave
Employee assistance program
Mental and financial wellness support
Flexible work arrangements: at home, in an office, or a combination of both