Applied AI/ML Scientist

Cerebras Systems builds the world's largest AI chip, 56 times larger than GPUs. Our novel wafer-scale architecture provides the AI compute power of dozens of GPUs on a single chip, with the programming simplicity of a single device. This approach allows Cerebras to deliver industry-leading training and inference speeds and empowers machine learning users to effortlessly run large-scale ML applications, without the hassle of managing hundreds of GPUs or TPUs. Cerebras' current customers include top model labs, global enterprises, and cutting-edge AI-native startups. OpenAI recently announced a multi-year partnership with Cerebras , to deploy 750 megawatts of scale, transforming key workloads with ultra high-speed inference. Thanks to the groundbreaking wafer-scale architecture, Cerebras Inference offers the fastest Generative AI inference solution in the world, 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. About The Role As an Applied AI Scientist in the FieldML team, you will be responsible for developing and customizing large language models and more broadly large-scale deep learning models to solve specific customer problems. You won't just advise; you will build. You will bridge the gap between state-of-the-art research and real-world applications by helping customers harness the power of the Cerebras Wafer-Scale Engine (WSE) for their AI initiatives. We are looking for experienced AI Scientists who are passionate about the "applied" side of machine learning - those who enjoy not just reading papers, but implementing, training, and scaling models to solve complex business and scientific problems. You will work on a diverse range of projects, from training bespoke models from scratch to fine-tuning and optimizing the latest Large Language Models (LLMs) for specific industry verticals, to designing and building components for custom agentic systems. The ideal candidate has experience in large model training and/or post-training, a deep understanding of training dynamics and model convergence, and expertise in data curation, combined with strong communication skills. Key Responsibilities Customer Use Case Discovery & Project Scoping Collaborate with customer stakeholders to identify the best approaches to their business problem with AI. Contribute to the technical scoping of engagements, including feasibility analysis, data quality/availability/readiness assessments, and the selection of optimal model architectures. Define project milestones, success metrics, and rigorous evaluation benchmarks to ensure the solution delivers measurable value to the customer’s business. Custom SOTA Models and AI Systems Development Architect and execute end-to-end training recipes for custom models, tailoring model architecture and training recipes to meet customer-specific performance and accuracy requirements. Design and implement sophisticated adaptation strategies, including continuous pre-training on private datasets, supervised fine-tuning (SFT), and post-training alignment via RLHF or DPO. Take full ownership of the training pipeline, from high-performance data preprocessing and tokenization to hyperparameter tuning and loss-curve analysis. Navigate the nuances of model convergence on specialized hardware, performing deep-dive analysis into loss dynamics and gradient stability. Scale training workloads across Cerebras clusters, ensuring efficient utilization of the hardware for multi-billion parameter models. Build and optimize the core components of agentic systems, focusing on tool-use capabilities, long-context reasoning, and multi-step planning. Technical Customer Leadership Serve as an AI/ML subject matter expert during technical deep-dives, translating customer requirements into precise training recipes. Build and ma