Physical AI Researcher - Multimodal Foundation Models

Archetype AI; San Mateo (CA); On-site; USD 140,000 - 210,000; Full time

Job summary

Archetype AI seeks an AI Researcher to design, train, and interpret large-scale models that learn from raw sensor streams. This role blends deep learning research, experimentation, and hands-on system building to advance real-world sensing capabilities.

You will work on representation learning, multimodal learning, and foundation models for diverse modalities including vibration, temperature, video, and other sensors.

Qualifications

• 8+ years of experience developing advanced ML/AI systems for real-world sensor data.
• Strong expertise in modern deep learning architectures, especially transformers and representation learning.
• Excellent programming skills in Python with PyTorch.

Responsibilities

• Design and train foundation models for sensor data, including multimodal architectures.
• Develop new approaches for representation learning and self-supervised learning from sensor streams.
• Adapt pretrained models to new datasets, modalities, environments, and use cases; balance data, capacity and performance.
• Collaborate with engineers and product teams to translate research into production systems.

Skills

Transformer models

Large-scale training

Python

PyTorch

Research & experimentation

Communication skills

Education

PhD or MS in CS/ML

Tools

PyTorch

Distributed training

ML systems

Job description

Archetype AI seeks an AI Researcher to design, train, and interpret large-scale models that learn from raw sensor streams. This role blends deep learning research, experimentation, and hands-on system building to advance real-world sensing capabilities.

You will work on representation learning, multimodal learning, and foundation models for diverse modalities including vibration, temperature, video, and other sensors.