ML Engineer: Multimodal Vision-Language

Spur; San Francisco (CA); On-site; USD 150,000 - 230,000; Full time

Benefits offered by this job

Health insurance

Stock options

Job summary

NomadicML is seeking a Machine Learning Engineer in San Francisco to advance foundation-model research and production engineering. You will train and fine-tune large-scale Vision-Language Models to reason about motion in real-world video, building multi-modal architectures that process video, language, and sensor data.

You will work closely with founders to deploy robust APIs and SDKs for enterprise use, publish research at top venues, and drive end-to-end experiments from data collection to

Qualifications

• Proficient in Python and PyTorch with large-scale ML workflows.
• Experience with foundation models, VLMs, or multi-modal learning.
• Ability to run end-to-end experiments and iterate quickly.
• Experience training or fine-tuning models on video or sensor data.
• Understanding of retrieval systems, embeddings, and GPU optimization.

Responsibilities

• Train and evaluate VLMs for motion understanding in autonomous-driving datasets.
• Design and scale GPU-accelerated pipelines for training, fine-tuning, and inference on multi-modal data.
• Build evaluation frameworks for spatiotemporal reasoning and narrative consistency.
• Develop curation loops that generate and refine datasets using models.
• Publish high-impact research while shipping customer-facing features.

Skills

Python

PyTorch

ML workflows

Experimentation

Tools

DeepSpeed

Hugging Face

Job description

NomadicML is seeking a Machine Learning Engineer in San Francisco to advance foundation-model research and production engineering. You will train and fine-tune large-scale Vision-Language Models to reason about motion in real-world video, building multi-modal architectures that process video, language, and sensor data.

You will work closely with founders to deploy robust APIs and SDKs for enterprise use, publish research at top venues, and drive end-to-end experiments from data collection to