Cerebras jobs

Staff Inference ML Runtime Engineer role at Cerebras, designing and implementing APIs, ML features, and tools for running generative AI models on custom hardware. Responsibilities include technical leadership, designing ML features (structured outputs, multimodal inference), optimizing latency/throughput, and maintaining serving infrastructure. Located in US and Canada offices; salary and equity not specified.

A Lead RTL Design Engineer role on Cerebras' Wafer Scale Engine (WSE) development, focusing on front-end chip design and integration. Responsibilities include RTL development, managing external ASIC vendors, collaborating with physical design/verification/software teams, and debugging silicon-level issues. Requires Master's in CS/EE (or equivalent), 8-15 years delivering complex RTL designs, experience with chip integration and third-party IP, proven silicon success, and ability to manage external vendors. Preferred skills include FPGA, high-speed IO, and networking stack (TCP/IP, RDMA, Ethernet). This is semiconductor/chip design, not software engineering. Position allows hybrid work but is based at Cerebras' Sunnyvale headquarters.

Staff Python/PyTorch Developer role on Cerebras' Inference Team focused on generative AI model optimization and compilation. Responsibilities include analyzing generative AI models, developing model representation frameworks using PyTorch, building frontend compiler infrastructure to ingest PyTorch models, and extending graph optimization tooling. Requires strong PyTorch internals expertise, C++, MLIR compiler stack knowledge, and experience with large language models. Located in UAE.

Principal Engineer (Reliability Tech Lead, IC) role at Cerebras Inference, owning reliability strategy and execution across distributed inference systems. Requires 7+ years backend/infrastructure/reliability engineering, expertise in distributed systems, and ability to design fault detection, failover, and incident response frameworks at scale. Located in US and Canada offices; salary and equity not specified.

Kernel Engineer developing high-performance ML and HPC kernels in low-level assembly and custom languages (CSL) for Cerebras hardware. Requires CS/engineering degree, C++/Python, and strong hardware architecture knowledge; located in India office.

A Staff SRE role building and leading the reliability function for Cerebras' AI inference service (powered by WSE). Responsibilities include defining reliability strategy and SLOs, architecting self-service platforms and GitOps-driven CD pipelines, automating toil (model releases, capacity provisioning, cluster upgrades), mentoring junior SREs, and driving platform adoption across teams. Requires 8+ years in SRE/infrastructure/platform engineering at scale (FAANG or hyperscaler), deep hands-on expertise with cloud/on-prem clusters, Kubernetes, Python/Go, observability, and incident response. The role is remote-eligible ('Remote Office') with no 24/7 on-call rotation, strong alignment with James' full-stack systems expertise and technical leadership background. Base compensation not specified.

Principal Engineer role for Inference Cloud Platform at Cerebras, owning the cloud layer behind Inference Service including multi-region architecture, reliability, and platform direction. Requires 10+ years distributed systems experience, deep cloud infrastructure expertise, and hands-on IC work on critical paths. Located in Sunnyvale HQ (on-site); salary and equity not specified.

This founding Product Manager role leads the Strategic Verticals team at Cerebras, embedding with strategic customers from AI-native startups to Fortune 500 enterprises. The PM owns customer outcomes from POC design through scaled deployment, advising on model selection, benchmarking performance, and translating customer insights into product roadmap requirements. The position combines product leadership, technical expertise, and GTM strategy, based in San Francisco.

Full Stack LLM Engineer on Cerebras' Inference Core Model Bringup team, rapidly bringing state-of-the-art open-source models (LLaMA, Qwen) onto Cerebras CSX systems. Responsibilities span model architecture translation, graph lowering, compiler optimizations, runtime integration, and performance tuning, with emphasis on debugging across the full stack. Requires Bachelor's+ in CS/Engineering, comfort with AI toolchain (Python, compiler IRs, performance profiling), deep learning framework experience (PyTorch/TensorFlow), C/C++ proficiency, and compiler development expertise (LLVM/MLIR). On-site in Toronto, Canada.

A 3D physical design engineer role focused on ASIC/SoC design, packaging, and 3D chip integration. Key responsibilities include designing and analyzing 3D integrated products, optimizing power/performance/area trade-offs, managing physical verification and IR/EM analysis, and collaborating with RTL teams. Requires 10+ years of physical design experience, expertise with Synopsys tools (ICV/Calibre), strong scripting (Tcl/Python), and specific 3D stacking/packaging experience. Compensation is $150k–$270k. Position is on-site at Cerebras' Sunnyvale headquarters.

Distributed Systems Cluster Security Software Engineering Lead position at Cerebras in Sunnyvale as security czar for large-scale AI clusters (100s of Wafer-Scale accelerators, 1000s of servers/networking ports). Core responsibilities include ensuring cluster security through first-principles engineering, developing security solutions for network/user access controls and multi-tenancy, building detection and response tools across the vertical stack, driving cross-functional collaboration, roadmap ownership, and building/managing an engineering team. Requires 3+ years in engineering leadership/management of distributed systems security, proven product delivery and customer deployment, excellent communication/collaboration, strong multi-tenancy and cluster networking background, and distributed systems software development expertise (Kubernetes preferred, bare-metal cluster management preferred). Salary: $140k–$240k.

Staff Software Engineer owning major architecture of Inference Cloud Platform (multi-region traffic, load balancing, reliability, global scale). Requires 8+ years software engineering with substantial distributed systems experience; on-site in Sunnyvale, CA.