Applied AI/ML Scientist, Cerebras Systems

Job Description

Cerebras Systems has announced a job opening for an Applied AI/ML Scientist at their FieldML in the United Arab Emirates. This job entails creating, training, fine-tuning, and optimizing huge language models and deep learning models for customer specific tasks. It will entail designing of agent-based models and scaling of models on special hardware.

Experience: Large model training and AI/ML

Qualification: Master’s or PhD in computer science or related field

Apply: Click Here

Responsibilities

• Determine the AI needs of the customer and select appropriate technology, architecture, milestones, and measures of evaluation.
• Create customized language and deep learning models via training, fine-tuning, and post-training techniques.
• Design data pipelines, optimize training processes, and perform model convergence, loss function dynamics, and gradient stability analysis.
• Scale multi-billion parameter models on the Cerebras clusters, optimising hardware and training efficiency.
• Work alongside the customer, research groups, and engineers to create AI solutions for specific needs.

Requirements

• Master’s or PhD degree in computer science, machine learning, or a related technical field.
• Expert understanding of transformers, MoEs, multimodal models, sequence architectures, scaling laws, and training dynamics.
• Proven experience training or fine-tuning large-scale models containing one billion or more parameters.
• Strong proficiency in Python and PyTorch, including distributed training and large-scale data processing frameworks.
• Strong communication and collaboration abilities for managing technical projects and explaining complex AI concepts.

Preferred Qualifications

• Hands-on experience in building agentic systems capable of tool use, reasoning in long context, and planning.
• Hands-on experience in continuous pre-training, supervised fine-tuning, RLHF, DPO, and other forms of model adaptations.
• Understanding of the process of data curation, tokenization, hyperparameter tuning, evaluation procedures, and post-training pipelines.
• Experience in optimizing performance, scalability, reliability, resource utilization, and efficiency of AI workloads.
• Experience collaborating in advanced AI initiatives involving customers, research, engineering, and product teams.