Runs the team day to day
In-house, owns the opening
Joined the team recently
From your own Gmail and LinkedIn.
About the role
This is a remote AI research engineering role at Tether, listed in Bengaluru, India, focused on compressing large language and vision-language models for edge devices such as smartphones. It suits researchers with deep experience in quantization, knowledge distillation and pruning, plus a background in multimodal transformer architectures. The work centers on cutting model size, latency and cost while keeping accuracy high, and on publishing the results.
What you’ll do
- Apply low-bit quantization to reduce model size and latency for generative models.
- Use knowledge distillation to move capabilities from large teacher models into smaller student models.
- Prune redundant parameters and attention heads without hurting task performance.
- Measure trade-offs between size, latency, memory and accuracy across compression methods, and propose improvements.
- Research and apply mixed-precision quantization and other advanced compression strategies.
- Track new compression research for multimodal and generative architectures.
- Build compression pipelines and set performance and fidelity metrics.
- Document methods and results so others can reproduce them.
- Write technical papers and submit findings to top-tier conferences.
What they’re looking for
- A degree in computer science or a related field, ideally a PhD in NLP, machine learning or a similar area
- A solid AI research record, ideally with publications at A* conferences
- Experience with PyTorch or an equivalent deep learning framework
- Hands-on experience with quantization-aware training and post-training quantization
- Research and hands-on experience with knowledge distillation
- Research and hands-on experience with model pruning
- A solid grasp of neural network architectures, including transformers, backpropagation, optimization and fine-tuning
Nice to have
- C++, especially for low-level quantization kernels or inference optimization
What’s on offer
- Remote work with a globally distributed team
Skills
Questions about this role
Is this role remote?
Yes, the listing marks it as remote.
Do I need a PhD?
A PhD is described as ideal, alongside a degree in computer science or a related field.
Is C++ required?
No, C++ is listed as a plus.
Which deep learning framework should I know?
PyTorch or an equivalent framework.