AI Research Engineer – Model Compression & Quantization

About the role

Join Tether’s AI research team to drive innovation in model compression and efficient deployment for advanced multimodal AI systems, including large language models and vision‑language models. You will focus on reducing model footprint and computational cost while preserving accuracy, enabling high‑performance AI on resource‑constrained edge devices.

Key responsibilities

• Apply low‑bit quantization to generative AI models to reduce size and latency while maintaining output quality.
• Leverage knowledge distillation to transfer capabilities from large teacher models to smaller student models for multimodal reasoning.
• Implement pruning techniques to remove redundant parameters and attention heads without sacrificing performance.
• Analyze trade‑offs between efficiency (size, latency, memory) and accuracy, proposing improvements based on empirical results.
• Research and apply mixed‑precision quantization and advanced compression strategies such as adaptive pruning schedules.
• Document methodologies, experiments, and results to support reproducibility and internal collaboration.
• Author technical papers and publish findings in top‑tier conferences (e.g., NeurIPS, ICML, ICLR, CVPR, ACL, AAAI).

Required profile

• Degree in Computer Science or related field; PhD in NLP, Machine Learning or similar is preferred.
• Strong research track record with publications in A* conferences.
• Hands‑on experience with model quantization (QAT and PTQ), knowledge distillation, and pruning.
• Deep understanding of neural network architectures, especially transformers for LLMs and VLMs.
• Experience with PyTorch or equivalent deep‑learning frameworks.

Required skills

• PyTorch (or equivalent deep‑learning framework)
• Model quantization techniques (Quantization‑Aware Training, Post‑Training Quantization)
• Knowledge distillation
• Model pruning
• Transformers and multimodal architectures (LLMs, VLMs)
• C++ (optional, for low‑level kernel implementation)