About the role
Tether, the company behind USDT, is hiring an AI Research Engineer to work on multi-modal vision-language models within its AI model team.
You'll run the full model lifecycle: data curation, post-training (supervised fine-tuning, distillation, RLHF), evaluation and optimisation for real-world deployment, including compression for resource-constrained environments. It is a research-driven role with strong engineering discipline, working on distributed GPU infrastructure in a small, high-caliber team.
Highlights
- Frontier multi-modal / vision-language research with production goals
- Post-training, distillation, RLHF, distributed training, model compression
- Fully remote worldwide, publish at top venues (NeurIPS, ICML, CVPR)
- MS/PhD preferred, open-source multimodal contributions a strong plus
This recap is dataskew's editorial summary, not the company's copy.