Staff AI/ML Software Engineer, Model Distillation & Fine-Tuning
Core
Lead the adaptation, fine-tuning, and distillation of foundation models for automotive edge deployment to understand driver intent, conversational context, and cabin visual state.
Role type
Staff AI/ML Software Engineer (Model Optimization & Distillation)
Builds
Compact multimodal models for vehicle edge hardware
Domain
Automotive AI / Edge Computing
Deliverable
production ML models
Required skills
PyTorch, parameter-efficient fine-tuning (LoRA/QLoRA), knowledge distillation, quantization-aware training, RLHF/DPO, dataset curation, model architecture strategy
Preferred skills
Quantized model shipping to specific hardware, large-scale training ecosystem (Hugging Face/DeepSpeed/Ray), conversational AI domain experience, open-source contributions to tuning libraries
Technologies
PyTorch, LoRA, QLoRA, RLHF, DPO, Hugging Face, DeepSpeed, Ray, Megatron
Responsibilities
Design knowledge distillation pipelines for edge deployment; Apply and scale parameter-efficient fine-tuning techniques; Build reinforcement learning flywheels with human-in-the-loop alignment; Curate and generate datasets for in-cabin interaction; Implement Quantization-Aware Training; Establish evaluation frameworks for fine-tuned models; Own base model strategy and architecture decisions
Seniority
Staff, individual contributor technical leader
