Software Engineer 6

About the Team

Netflix Cloud is the infrastructure every Netflix service runs on. Within it, the Infrastructure Management org builds the platforms engineers use to provision, configure, and operate it - safely, at scale, across every team at the company. We build and run Netflix’s infrastructure-as-code product - the front door to Netflix Cloud - integrated with Spinnaker and our SDLC. We also own the durable execution platform powering Tier‑0 control planes and increasingly serving as the backbone for AI/agentic workloads. And we’re developing agent sandbox primitives so agent execution can scale safely across every compute environment as agentic development outpaces governance. We're looking for engineers who think like startup founders, partner deeply across the company, and treat the platform as a product with real users and real stakes.

What You Will Do

• Scale Netflix’s developer platforms from early adoption to broad, exponential usage by making high-leverage product decisions and partnering across all engineering orgs

• Own the end-to-end Netflix Cloud developer experience across the SDLC, from discovery and configuration through deploy, operate, and debug

• Build and run Tier‑0 control planes for core infrastructure workflows, with direct accountability for reliability, scalability, and cost

• Define how AI agents safely interact with Netflix infrastructure by building sandboxing, identity, and governance primitives and shared human/agent interfaces

• Set the boundary between self-serve and guardrails by choosing platform defaults, provider guarantees, and where the platform stays out of the way

What We Are Looking For

• Has built or scaled a developer platform in production - not contributed to one, owned one

• High product leverage: consistently makes strong product decisions, builds backward from customer goals and adoption constraints, not forwards from tech

• Comfortable in ambiguity: sets direction with minimal structure, iterates via PoCs, converges quickly, solutions are simpler than the problems

• Comfortable in a multi-language, multi-service codebase (Java/Kotlin/Go or equivalent mix); ramps fast on unfamiliar systems

• Strong distributed systems fundamentals: race conditions, signal synchronization, control plane design; has failures to prove it