Site Reliability Engineer, Provider Operations

OpenRouter routes close to a billion LLM requests a day across 80+ providers and thousands of endpoints. This is its first AI inference SRE, on the Provider Operations team, owning the health of that provider supply.

You would build monitoring for every provider and endpoint covering latency, throughput, errors, uptime and output correctness, set SLOs per provider tier, and work with the routing team to detect degraded endpoints and route around them automatically. You would own on-call for provider incidents, produce scorecards providers act on, and build canaries and evals that catch silent regressions such as quantisation changes, broken tool calling and truncated streams.

Pay is $245,000 to $295,000. Remote, US.