Leads the team you would join
Signs off on engineering hires
Owns the opening
From your own Gmail and LinkedIn.
About the role
This is a full-time, 100% remote AI research engineering role on Tether's AI model team, open to candidates worldwide. It suits researchers with strong engineering discipline who want to train and tune vision-language models for real-world deployment. Tether is a fintech group whose offerings include the USDT stablecoin, digital asset tokenization services, and energy and data businesses. The team is small and works across time zones, focusing on models that are both capable and efficient to run at scale.
What you’ll do
- Carry out end-to-end research and engineering on vision-language models, covering training, evaluation, and optimization.
- Design post-training pipelines using supervised fine-tuning, knowledge distillation, and reinforcement learning from human feedback.
- Curate, filter, and balance multimodal datasets for domain-specific tasks.
- Make models smaller and faster for constrained environments using compression and optimization methods.
- Build evaluation frameworks and benchmarks that measure quality, robustness, and real-task success.
- Scale training workflows across distributed GPU infrastructure and remove pipeline bottlenecks.
- Use open-source models, datasets, and tooling where it speeds up development.
- Track multimodal research and turn useful findings into practical improvements.
- Publish findings at leading AI conferences and journals where appropriate.
What they’re looking for
- A research-driven mindset combined with strong engineering discipline.
- The judgment to choose the right technique for a problem, implement it rigorously, and measure its impact clearly.
- Excellent English communication skills.
What’s on offer
- 100% remote, worldwide, with flexible working hours across time zones.
- Competitive compensation in USD and/or crypto.
- Exposure to stablecoin infrastructure, AI, and Bitcoin mining.
- A globally distributed team across 20+ countries.
Skills
Questions about this role
Is this role remote?
Yes. It is 100% remote and open worldwide.
What kind of work is involved?
Training, evaluating, and optimizing vision-language models, including post-training pipelines and dataset curation.
How is compensation paid?
Competitive compensation in USD and/or crypto.
Is English required?
Excellent English communication skills are asked for.