AI Inference Core - Software Integration Engineer

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 are looking for a Software Integration Engineer to join the AI Inference Core team at Cerebras. This team sits at the intersection of AI infrastructure, distributed systems, compilers, runtimes, kernels, and hardware/software co-design. The Innovation Engine for Inference Core — turning ideas into reality. You will help take ambitious ideas from concept to working reality across the Cerebras inference stack. You will integrate and validate cross-component projects of high complexity, often on accelerated timelines, and work directly with engineers across AI, runtime, compiler, kernel, systems, and hardware teams. A SPECIAL TASK FORCE, NOT A TYPICAL ENGINEERING ROLE - Zero-to-one mission: Take incomplete ideas and early prototypes all the way to working, validated capabilities. - Cross-stack complexity: Move across AI frameworks, runtime, compiler, kernels, distributed systems, infrastructure, and hardware—not just one component or codebase. - Accelerated and dynamic cadence: Expect focused daily syncs, rapidly changing priorities, and periods of intense integration work. This is not a role for engineers seeking strictly regular, predictable work hours. - Comfortable on the hot seat: Take ownership when the path is unclear, make sound decisions with incomplete information, and stay effective when timelines are tight. - Bias for action: At critical moments, the mindset is “No Process, No Documentation, Just Get Stuff Done!!!” Cut through unnecessary ceremony, deliver the result, and then turn what you learned into better automation, diagnostics, documentation, and repeatable practices. - Team multiplier: Always push the work forward while raising the pace, clarity, and effectiveness of the people around you. We want people who bring the team with them—not lone heroes. We prefer candidates with experience in software/hardware co-design or other complex systems, but that experience is not required. We welcome exceptional junior engineers who are eager to learn, do not shy away from ambiguity or hard work, and can demonstrate strong fundamentals, curiosity, ownership, and persistence. WHAT YOU WILL DO - Turn new inference ideas and features into integrated, working capabilities across the Cerebras platform. - Take high-value projects from zero to one—from an incomplete idea or prototype to a working, validated capability. - Integrate and validate software components spanning AI frameworks, runtime, compiler, kernels, distributed systems, and hardware. - Drive cross-component projects of high complexity from problem definition through integration, validation, and release readiness. - Collaborate closely with software and hardware engineers to make co-design trade-offs and resolve system-level issues. - Investigate and debug difficult failures across large-scale AI workloads, distributed software, infrastructure, and hardware boundaries. - Work effectively on accelerated timelines while keeping technical risks, dependencies, and decisions visible. - Participate in focused daily syncs, manage rapidly changing priorities, and drive ambiguous situations toward concrete outcomes. - Identify bottlenecks, failure modes, edge cases, and integration gaps that affect inference correctness, performance, or delivery. - Capture the essential lessons from urgent work so the next integration is faster and less chaotic. - Create momentum beyond your own work by helping teammates move faster, make better decisions, and close difficult problems together. MINIMUM SKILLS & QUALIFICATIONS - Strong software-engineering fundamentals and programming ability in Python, C++, Go, or a similar language. - Experience with software/hardware co-design, hardware accelerators, compilers, kernels, runtimes, or low-level systems. - Experience debugging complex systems, distributed software environments, or large-scale compute clusters. - Experience with AI infrastructure, model deployment, LLMs, or multimodal workloads. - Exposure to performance debugging, profiling, observability, or failure analysis. - Familiarity with microservices, containers, cluster orchestration, cloud infrastructure, or high-performance computing. - Experience—through professional work, internships, research, academic projects, open source, or equivalent hands-on work—building or debugging software systems. - Clear communication and strong collaboration across disciplines and levels of experience. PREFERRED SKILLS - Experience in a startup or similarly fast-moving, resource-constrained engineering environment. - Demonstrated experience taking a project from zero to one: turning an ambiguous problem, early idea, or prototype into a working and reliable capability. - Demonstrated ability to break down ambiguous technical problems, form hypotheses, gather evidence, and drive issues to resolution. - Curiosity about how complex systems behave across component boundaries. - Willingness to read unfamiliar code, learn new layers of the stack, and take ownership beyond a narrowly defined area. - Ability to work hard and stay effective during periods of uncertainty, rapid change, and accelerated delivery. - Track record of driving cross-team projects from an incomplete idea to a rel