AI Research Engineer (Pre-training - LLM & Multi-Modal)

Remotedxb; Dubai; On-site; AED 2,500,000 - 4,000,000; Full time

Job summary

Tether in the United Arab Emirates seeks an AI Research Engineer to lead foundational pre-training for large language and multimodal models on distributed GPU clusters. You will design architectures, tokenizers, and alignment layers, while curating massive data pipelines and running rigorous experiments.

The role requires deep knowledge of transformer models, hands-on PyTorch and Hugging Face experience, and a track record in publications.

Qualifications

• Degree in Computer Science or related field (PhD preferred).
• Proven AI R&D track record with publications in top conferences.
• Hands-on experience with large-scale LLM or multimodal pre-training on distributed GPUs.
• Proficiency with PyTorch and Hugging Face libraries.
• Experience with large-scale distributed training frameworks.

Responsibilities

• Conduct foundational pre-training for LLMs and multimodal models on large GPU clusters.
• Design architectures, tokenizers, and cross-modal alignment layers.
• Source, filter, and curate massive datasets and establish robust data pipelines.
• Execute experiments, analyze results, and refine training methodologies.
• Investigate bottlenecks in model efficiency and multi-modal alignment stability.
• Contribute to distributed training systems for hardware efficiency.

Skills

AI R&D

Publications

Distributed training

Dataset pipelines

Education

Degree in Computer Science or related field

PhD in NLP or Machine Learning preferred

Tools

PyTorch

Hugging Face

Distributed training frameworks

Job description

Responsibilities

• Conduct foundational pre-training for LLMs and Multi-Modal models on large, distributed servers with thousands of NVIDIA GPUs
• Design, prototype, and scale innovative architectures, tokenizers, and cross-modal alignment layers
• Source, filter, and curate massive-scale textual and multi-modal datasets and establish robust data pipelines
• Execute experiments, analyze results, and refine training methodologies for optimal performance
• Investigate, debug, and eliminate bottlenecks in model efficiency and multi-modal alignment stability
• Contribute to the advancement of distributed training systems for hardware efficiency

Requirements

• Degree in Computer Science or related field (PhD in NLP or Machine Learning preferred)
• Track record in AI R&D with publications in A* conferences
• Hands-on experience with large-scale LLM or Multi-Modal pre-training on distributed GPU clusters
• Practical experience with large-scale distributed training frameworks and tools
• Deep knowledge of state-of-the-art transformer and non-transformer modifications
• Strong expertise in PyTorch and Hugging Face libraries

About the Company

Tether is a leader in digital finance, pioneering solutions that empower businesses to integrate reserve-backed tokens across blockchains. From the world's most trusted stablecoin, USDT, to cutting-edge AI and peer-to-peer technology, Tether is shaping the future of the global financial revolution.