Software Engineer, Lambda

• 3+ years of experience building production systems in Python or Go
• Are comfortable with Linux systems and understand systems-level operations
• Understanding of control plane design patterns, including reconciliation loops, desired vs. actual state, and event-driven orchestration
• Familiarity with scheduling, bin packing, and resource allocation strategies at scale
• Think clearly about system architecture, automation, and the gap between desired and actual state

Desirable

• Lambda is hiring a Software Engineer to join our Fleet Orchestration Team, where we’re building the systems that automate and coordinate the lifecycle of thousands of GPU hosts across Lambda’s global infrastructure
• Our systems act as the control and data planes for our physical fleet, enabling dynamic allocation of machines between products, full automation of provisioning workflows, and accurate tracking of infrastructure state
• You’ll be responsible for developing services and tooling that convert product needs into infrastructure actions—handling everything from initial host onboarding to reconfiguration and retirement
• This is an opportunity to build core backend systems that make Lambda’s infrastructure more flexible, observable, and reliable
• Develop Control Plane Services: Design and build backend software that manages physical host allocation, deployment workflows, and state reconciliation
• Automate Host Lifecycle Management: Implement systems that orchestrate provisioning, OS installation, firmware updates, and decommissioning of GPU machines
• Ensure Infrastructure Accuracy: Build tools to validate and maintain consistency between intended and actual system state across large-scale hardware environments
• Enable Dynamic Resource Allocation: Create mechanisms that support shifting machine usage across internal and customer-facing products based on real-time demand
• Collaborate with Cross-Functional Teams: Partner with teams across infrastructure, security, and product engineering to deliver scalable, maintainable solutions