Machine Learning Engineer, Multimodal Perception and Authentication

OpenAI; United States; Hybrid; USD 140,000 - 210,000; Full time

Benefits offered by this job

Relocation assistance

Job summary

OpenAI's Future of Computing Research team in San Francisco is seeking a machine learning engineer to help shape how AI systems understand the physical world and the people in it.

You will focus on multimodal perception and authentication, bringing together signals from cameras, microphones, and other sensors, and working with hardware, firmware, software, and product teams to turn research into real-world systems.

Qualifications

• Strong background in computer vision, audio or speech machine learning, multimodal learning, or sensing.
• Experience developing specialized ML models, larger multimodal models, or both.
• Experience bringing research ideas into practical systems, prototypes, or products.
• Know how to design experiments, build evaluations, and investigate model behavior.

Responsibilities

• Research and develop multimodal perception and authentication methods across visual, audio, and other sensing signals.
• Explore how specialized perception models and larger multimodal models can work together.
• Design data, training, and evaluation approaches that improve performance in real-world conditions.
• Study model behavior, robustness, and failure modes across sensing, data, and deployment environments.
• Integrate and validate new capabilities in real-time or resource-constrained systems.
• Work with hardware, firmware, software, and product teams to turn research into working systems.

Skills

Computer vision

Audio or speech ML

Multimodal learning

Sensing

Job description

About the Team

The Future of Computing Research team is an applied research team within OpenAI’s Consumer Devices group. We study how AI systems perceive people and their surroundings, and we turn that research into capabilities for future products. Our work spans machine learning, sensing, and hardware, with a focus on building systems that work beyond controlled environments.

About the Role

We’re looking for a machine learning engineer to help shape how future AI systems understand the physical world and the people in it. The role focuses on multimodal perception and authentication, bringing together signals from cameras, microphones, and other sensors. You’ll work with specialized perception models and larger multimodal models, and partner with hardware, firmware, software, and product teams to bring new research into real-world systems.

This role is based in San Francisco. We work in the office three days per week and offer relocation assistance.

In this role, you will:

• Research and develop multimodal perception and authentication methods across visual, audio, and other sensing signals.
• Explore how specialized perception models and larger multimodal models can work together.
• Design data, training, and evaluation approaches that improve performance in real-world conditions.
• Study model behavior, robustness, and failure modes across sensing, data, and deployment environments.
• Integrate and validate new capabilities in real-time or resource-constrained systems.
• Work with hardware, firmware, software, and product teams to turn research into working systems.

You might thrive in this role if you:

• Have a strong background in computer vision, audio or speech machine learning, multimodal learning, or sensing.
• Have experience developing specialized machine learning models, larger multimodal models, or both.
• Have brought research ideas into practical systems, prototypes, or products.
• Know how to design experiments, build evaluations, and investigate model behavior.
• Have wo