Senior AI scientist
Core
Design, train, and scale large-scale transformer and hybrid foundation models for conversational AI platforms using multimodal LLMs on Edge and cloud.
Role type
Senior IC AI scientist (foundation models)
Builds
Next-generation conversational AI models for automotive voice assistants
Domain
Automotive + Large Language Models
Deliverable
production ML models
Required skills
Deep learning fundamentals, transformer architecture design, optimization dynamics, scaling laws, distributed training, loss function design
Preferred skills
Experience with MoE, RLHF/DPO alignment, mixed precision training
Technologies
FSDP, ZeRO-3, Tensor parallelism, Pipeline parallelism, AdamW, Lion, Adafactor, bf16, fp8
Responsibilities
Design and modify transformer architectures (Attention variants, RoPE, ALiBi, GQA, MoE); Diagnose and resolve training instabilities; Navigate scaling tradeoffs across data, compute, and architecture; Define technical direction for next-generation models; Apply and validate scaling laws; Design and experiment with loss functions (Next-token prediction, Contrastive, RLHF, DPO, GRPO); Design and execute large-scale distributed training.
Seniority
Senior, hands-on IC