Research Engineer-Model Efficiency
Core
Implement and adapt model-efficiency methods (quantization, sparsity, pruning, speculative decoding, MTP, serving-time attention, KV-cache) on Bitdeer's models and hardware to improve performance and quality.
Role type
Research Engineer (Model Efficiency)
Builds
Optimized inference models and serving systems for digital asset technologies
Domain
AI/ML, Model Efficiency, Hardware-Aware Optimization
Deliverable
production ML models
Required skills
Python, PyTorch, Transformer architecture internals, Quantization, Sparsity, Pruning, Speculative decoding, KV-cache management, Rigorous evaluation design
Preferred skills
Open-source contributions, Academic publications, Algorithm competitions, Deep project experience
Technologies
PyTorch, Python
Responsibilities
Implement published efficiency methods, Develop new optimizations, Build rigorous evaluations, Present defined projects with real results
Seniority
Intern (Undergraduate/Master's/PhD candidate)