AI Research Engineer (Kernel & Inference Optimization)
Responsibilities:
Design and deploy state-of-the-art model serving architectures for high throughput and low latency
Build and monitor controlled inference tests in simulated and live production environments
Identify and prepare high-quality test datasets and simulation scenarios for low-resource devices
Analyze computational efficiency and diagnose bottlenecks in the serving pipeline
Integrate optimized serving and inference frameworks into production pipelines for edge and on-device applications
Requirements:
Degree in Computer Science or related field; PhD in NLP or Machine Learning preferred
Proficiency in Metal Shading Language (MSL) and writing custom compute shaders
Proven experience in low-level kernel optimizations and inference optimization on mobile devices
Deep understanding of modern model serving architectures and inference optimization techniques
Expertise in writing GPU kernels for mobile devices and deploying end-to-end inference pipelines
Ability to apply empirical research to overcome latency and memory constraints
Experience with Distributed Inference Systems (Tensor, Pipeline, and Expert Parallelism)
Deep understanding of Diffusion Models and Vision Transformers
Knowledge of Pruning, Quantization, Flash attention, KV Cache, and Speculative Decoding
About the Company:
Tether is a global fintech leader pioneering a financial revolution through reserve-backed tokens and stablecoins like USDT, while also expanding into sustainable energy, AI data solutions, and digital education.