Principal Engineer, Inference Cloud
Cerebras · Headquarters/Sunnyvale Office
Cerebras Systems builds the world's largest AI chip, 56 times larger than GPUs. This architecture allows Cerebras to deliver industry-leading training and inference speeds; over 10 times faster than GPU-based hyperscale cloud inference services. This order of magnitude increase in speed is transforming the user experience of AI applications, unlocking real-time iteration and increasing intelligence via additional agentic computation. Cerebras works with the leading model labs, global enterprises, and cutting-edge AI-native startups. OpenAI recently announced a multi-year partnership https://openai.com/index/cerebras-partnership/ with Cerebras, to deploy 750 megawatts of scale, transforming key workloads with ultra high-speed inference. About The Role We're hiring a Principal Engineer for our Inference Cloud Platform. This team owns the cloud layer behind our Inference Service, including availability, latency, reliability, and multi-region scale. This is one of the most senior IC roles on the team, for someone who can identify the highest-leverage platform problems, set direction across multiple teams, define long-term architecture, and write production code on critical paths. Many of the key decisions are ambiguous at the outset; you’ll need to frame the problem, make tradeoffs, and drive execution without a clear spec. The scope includes multi-region traffic architecture, graceful degradation under bursty AI workloads, high-QPS performance, and the operating model for a platform that needs to remain fast and available under changing demand. You'll partner closely with ML, Product and Infrastructure teams. Responsibilities - Problem Definition & Prioritization. Identify the most important technical problems for the platform, often before there's a clear ask. Make explicit tradeoff decisions about what the platform will and won't support, with reasoning that holds up under scrutiny from senior engineering leadership. - Platform Direction. Set the lo