AI Research Engineer (Pre-training)

TL;DR

AI Research Engineer (Pre-training): Advancing large-scale pre-training systems for state-of-the-art AI models with an accent on model architectures, training efficiency, and scalability across massive GPU clusters. Focus on designing novel transformer and non-transformer models, optimizing distributed training pipelines, and resolving bottlenecks in high-performance AI workloads.

Location: Remote (Mexico)

Company

Frontier AI research organization focused on next-generation model development at massive scale.

What you will do

• Conduct large-scale pre-training of AI models on distributed GPU clusters for scalability and performance
• Design, prototype, and optimize novel model architectures including transformers
• Run experiments, analyze results, and refine methodologies to boost training efficiency and model quality
• Identify and resolve bottlenecks in training systems, data pipelines, and model performance
• Improve distributed training infrastructure for next-generation AI workloads
• Collaborate with researchers and engineers to productionize experimental ideas

Requirements

• PhD or strong background in Computer Science, Machine Learning, NLP, or related fields (preferred)
• Hands-on experience with large-scale LLM pre-training on distributed GPU infrastructure (thousands of GPUs)
• Strong understanding of transformer architectures and advanced model design
• Experience with distributed training frameworks and large-scale AI systems
• Proficiency in PyTorch and Hugging Face ecosystem
• Strong skills in debugging, optimizing model and system performance, designing experiments, and iterating on research

Culture & Benefits

• Remote-friendly and globally distributed work environment
• Access to high-performance computing infrastructure and large GPU clusters
• Collaborative setting with top-tier AI researchers and engineers
• High autonomy in research direction and experimentation
• Competitive compensation aligned with AI research market standards
• Exposure to state-of-the-art AI systems and professional growth in innovation-driven ecosystem