Remote AI Research Engineer: Multimodal Model Compression

Tether Ltd.; Dubai; Remote; AED 661,000 - 881,000; Full time

Job summary

Tether Ltd. is seeking a senior AI research scientist to advance model compression for multimodal systems, including LLMs and VLMs, optimizing for edge devices while preserving fidelity and efficiency.

You will prototype quantization, distillation, and pruning pipelines, set performance metrics, and publish results at leading AI conferences. This role supports global, remote collaboration across teams and stakeholders.

Qualifications

• Degree in Computer Science or related field; PhD preferred in NLP/ML with strong AI R&D track record.
• Experience with PyTorch or equivalent frameworks.
• Hands-on with model quantization including QAT and PTQ.
• Research and hands-on with distillation for compressing large models.
• Experience with model pruning for efficient models.
• Solid understanding of transformers and multimodal architectures.
• Familiarity with C++ is a plus for low-level kernels.

Responsibilities

• Apply low-bit quantization to reduce model size and latency while preserving accuracy.
• Use knowledge distillation to transfer capabilities from large to small models.
• Implement pruning to remove redundant parameters and heads.
• Analyze trade-offs between size, latency, memory and accuracy.
• Research mixed-precision quantization and adaptive pruning schedules.
• Stay updated with state-of-the-art in multimodal and generative AI.
• Document methodologies, experiments, and results for reproducibility.
• Publish findings in top conferences (NeurIPS, ICML, ICLR, CVPR, ACL, AAAI).

Skills

Compression techniques

Multimodal AI

Quantization

Knowledge distillation

Model pruning

Edge deployment

LLMs/VLMs

Performance metrics

Education

PhD in NLP or Machine Learning

Tools

PyTorch

C++ (plus)

Job description

Tether Ltd. is seeking a senior AI research scientist to advance model compression for multimodal systems, including LLMs and VLMs, optimizing for edge devices while preserving fidelity and efficiency.

You will prototype quantization, distillation, and pruning pipelines, set performance metrics, and publish results at leading AI conferences. This role supports global, remote collaboration across teams and stakeholders.