Senior Multimodal AI Scientist - Scalable LLM/VLM Systems

Adobe; San Jose (CA); On-site; USD 216,000 - 313,000; Full time

Job summary

Adobe Firefly’s ASML group seeks an Applied Scientist to define technical strategy for multimodal data intelligence systems and to architect distributed LLM/VLM inference platforms. You will work across modeling, infrastructure, data, evaluation, and product teams to deliver scalable AI capabilities at scale.

You will influence the next generation of Firefly models by improving data quality, model efficiency, and scalable AI infrastructure used by millions of creators worldwide.

Qualifications

• PhD or MS in CS, ML or related field with significant industry experience.
• Deep expertise in LLMs/VLMs or multimodal models with production serving experience.
• Strong background in distributed systems and cloud-native ML infra.
• Proven experience optimizing large-scale inference workloads and GPU utilization.
• Hands-on in building high-throughput data pipelines for multimodal tasks.
• Proficiency with modern inference frameworks such as vLLM, TensorRT-LLM, Triton.

Responsibilities

• Architect and optimize distributed multimodal inference pipelines for image, video, and audio tasks.
• Drive LLM/VLM inference optimization, batching, quantization, and model serving.
• Build scalable data generation workflows using multimodal foundation models.
• Lead strategy for automated data annotation, filtering, and metadata enrichment.
• Design distributed processing systems for petabyte-scale datasets.
• Collaborate with research to productionize new capabilities and ensure reliability.

Skills

LLMs & VLMs

Distributed systems

Python

PyTorch

Cloud-native ML infra

Model serving

Multimodal pipelines

GPU utilization

Communication

Education

PhD or MS in Computer Science, ML, or related field

Tools

Ray/Spark/Dask

Kubernetes

TensorRT-LLM

Triton Inference Server

Job description

Adobe Firefly’s ASML group seeks an Applied Scientist to define technical strategy for multimodal data intelligence systems and to architect distributed LLM/VLM inference platforms. You will work across modeling, infrastructure, data, evaluation, and product teams to deliver scalable AI capabilities at scale.

You will influence the next generation of Firefly models by improving data quality, model efficiency, and scalable AI infrastructure used by millions of creators worldwide.