Remote opportunity at
Tether.ioAI Research Engineer (Model Compression & Quantization)
Tether is hiring a remote, full-time AI Research Engineer focused on model compression and quantization. The role sits within the company's AI research team. Tether describes…
Career Tools
About This Role
Tether is hiring a remote, full-time AI Research Engineer focused on model compression and quantization. The role sits within the company's AI research team. Tether describes itself as a digital finance company whose products include the USDT stablecoin, and this position is within its AI work. The posting does not state a salary. The engineer would reduce the size and computational cost of large language models and vision-language models while…
Job Description
Tether is hiring a remote, full-time AI Research Engineer focused on model compression and quantization. The role sits within the company's AI research team. Tether describes itself as a digital finance company whose products include the USDT stablecoin, and this position is within its AI work. The posting does not state a salary.
The engineer would reduce the size and computational cost of large language models and vision-language models while keeping accuracy high, so that multimodal systems can run on resource-limited edge devices such as smartphones. The work includes applying quantization, knowledge distillation, and pruning, then measuring the trade-offs between footprint, speed, memory use, and accuracy.
Other responsibilities include building compression pipelines, setting metrics for speed and output fidelity, and resolving slowdowns in production inference. The engineer would also track emerging compression research for multimodal and generative models, record methods and results so they can be reproduced, and contribute to papers for top-tier conferences such as NeurIPS, ICML, ICLR, CVPR, ACL, and AAAI.
Applicants need a computer science degree or a closely related discipline. Doctoral study in NLP, machine learning, or an adjacent area, with strong publications in leading venues, is ideal. Hands-on experience with both quantization-aware training and post-training quantization is required, along with knowledge distillation, pruning, PyTorch, and a solid understanding of transformer architectures. C++ is a welcome extra, and excellent English communication is expected.
Responsibilities
- Use low-bit quantization to shrink generative models and speed up inference
- Use distillation to pass skills from large teacher networks down to compact student networks
- Prune redundant parameters and attention heads without hurting task performance
- Compare efficiency and accuracy across methods and propose improvements from results
- Record methods, experiments, and results for reproducibility and stakeholder updates
Requirements
- Computer science or closely related degree
- Hands-on quantization experience, including quantization-aware training and post-training quantization
- Hands-on knowledge distillation and pruning for compressing large models
- Experience with PyTorch or equivalent deep learning frameworks
- Solid understanding of transformer architectures, backpropagation, optimization, and fine-tuning
- Excellent English communication skills
Qualifications
- Doctoral degree in NLP, machine learning, or an adjacent field (ideal)
- Publication record in leading AI conferences
- C++ experience, useful for writing performance-critical inference code (a plus)
Core Skills
Frequently Asked Questions
Answers are based only on the employer’s listing; where it doesn’t say, neither do we.
Is the AI Research Engineer job remote?
Yes, the role is listed as remote and full-time.
What technical skills are required?
The posting requires hands-on quantization (both QAT and PTQ), knowledge distillation, pruning, and PyTorch experience.
What education is needed?
A computer science or closely related degree is required, and a PhD with strong publications is described as ideal.
What is the salary?
The posting does not specify a salary or benefits.
Sample Interview Questions
AI-generated questions tailored to this specific role — a preview of the full practice set.