AI Research Engineer (Model Compression & Quantization)

Stack

Responsibilities

• ▹Apply and advance compression techniques such as quantization, knowledge distillation, and pruning for multimodal architectures integrating text, images, and audio
• ▹Develop, test, and implement novel compression strategies balancing model size, latency, throughput, and accuracy
• ▹Build robust compression pipelines
• ▹Establish performance and fidelity metrics
• ▹Address bottlenecks in production inference
• ▹Deliver scalable, low-memory, low-latency AI systems on edge devices (e.g. smartphones)

Requirements

• ▹Deep expertise in model compression methods
• ▹Strong background in multimodal model architectures
• ▹Excellent English communication skills

Soft skills

About the company

Tether is pioneering a global financial revolution: it operates the world's most trusted stablecoin, USDT, alongside digital asset tokenization services, sustainable Bitcoin mining energy solutions (Tether Power), AI and peer-to-peer technology (Tether Data, including the KEET app), digital education programs (Tether Education), and forward-looking innovation initiatives (Tether Evolution). The team works fully remotely from around the world.