Edge AI Research Engineer: Multimodal Model Compression

Tether; Dublin; On-site; EUR 70,000 - 90,000; Full time

Job summary

Tether is seeking a member of their AI research team to drive innovation in model compression for advanced multimodal AI systems, focusing on efficient deployment of large language models and vision-language models. The role demands expertise in experimental methods such as quantization, distillation, and pruning.

Candidates should possess a PhD in NLP or Machine Learning and have experience with model compression techniques. This position will enhance the performance of AI systems running on resource-constrained devices in Ireland.

Qualifications

• PhD in NLP, Machine Learning, or a related field, complemented by a solid track record in AI R&D.
• Experience with PyTorch deep learning frameworks or equivalent.
• Hands-on experience with model quantization (QAT and PTQ).
• Research experience with knowledge distillation and model pruning.

Responsibilities

• Apply low-bit quantization to reduce model size and latency for generative AI models.
• Leverage knowledge distillation to enable efficient multimodal reasoning.
• Implement pruning techniques for reducing computational overhead.
• Research and apply advanced compression strategies to optimize accuracy-performance balance.
• Document methodologies and publish findings in top-tier conferences.

Skills

Model Compression

Quantization

Knowledge Distillation

Model Pruning

PyTorch

Neural Network Architectures

Education

PhD in NLP, Machine Learning, or related field

Degree in Computer Science or related field

Job description

Tether is seeking a member of their AI research team to drive innovation in model compression for advanced multimodal AI systems, focusing on efficient deployment of large language models and vision-language models. The role demands expertise in experimental methods such as quantization, distillation, and pruning.

Candidates should possess a PhD in NLP or Machine Learning and have experience with model compression techniques. This position will enhance the performance of AI systems running on resource-constrained devices in Ireland.