LLM/VLM AI Research Engineer

ASML; San Jose (CA); On-site; USD 163,000 - 239,000; Full time

Job summary

ASML in San Jose is seeking AI Research and Development Engineers to advance LLM/VLM research for semiconductor inspection. You will own end-to-end development from data prep to deployment, build Agentic AI systems, and collaborate with cross-functional teams to translate research into prototypes.

The role emphasizes hands-on training, evaluation, and optimization of models, with opportunities to influence future products and apply multimodal AI to complex engineering challenges.

Qualifications

• Master’s or PhD in CS/AI/ML or related field.
• Hands-on experience training/fine-tuning LLMs or VLMs end-to-end.
• Proficiency with Python and PyTorch.
• Understanding transformer models, tokenization, and training dynamics.
• Experience with distributed multi-GPU training and debugging.

Responsibilities

• Own end-to-end LLM/VLM development workflows from data prep to deployment.
• Design model adaptation strategies like LoRA or fine-tuning.
• Build and maintain training and evaluation datasets with quality controls.
• Establish evaluation benchmarks for quality, latency, and cost.
• Develop Agentic AI solutions integrating tools and workflows.
• Apply multimodal AI techniques to engineering applications.
• Collaborate with cross-functional teams to define research objectives.
• Transition validated research into reusable platforms.

Skills

LLM training

VLM training

PyTorch

Python

Distributed training

Experimental design

Communication

Education

Master's/PhD in CS/AI/ML

Tools

PyTorch

Multi-GPU systems

Model development tooling

Job description

ASML in San Jose is seeking AI Research and Development Engineers to advance LLM/VLM research for semiconductor inspection. You will own end-to-end development from data prep to deployment, build Agentic AI systems, and collaborate with cross-functional teams to translate research into prototypes.

The role emphasizes hands-on training, evaluation, and optimization of models, with opportunities to influence future products and apply multimodal AI to complex engineering challenges.