Machine Learning Engineer – Real-Time Multimodal Perception

New Today

About the Role OpenAI seeks a Machine Learning Engineer to build multimodal ML systems that deliver secure, low‑friction user authentication and intelligent device perception. You will work at the intersection of modeling and systems engineering, architecting data pipelines and defining durable feature interfaces for video, audio, and future signals. You will build perception and decision pipelines and harden everything for deployment in real‑world environments. Key Responsibilities Develop multimodal authentication and identity‑confidence models across video and audio, with room to incorporate additional signals as they become available. Drive robustness across diverse conditions.
Architect data systems and scalable data‑augmentation and simulation pipelines that expand edge‑case coverage and increase model value.
Build and tune production inference and decision stacks, and use telemetry to improve reliability in the field.
Develop failure‑analysis systems that detect drift, false accepts and false rejects, and calibration issues, enabling rapid iteration loops.
Partner closely with hardware, firmware, and research teams to integrate sensors, shape model design, and take systems to production.
Ideal Candidate Has shipped ML systems in production where reliability mattered (e.g., safety, access control, payments).
Brings experience with authentication, biometrics, or access‑control machine learning.
Strong background in computer vision, audio ML, and multimodal fusion.
Proficient in C++ and Python with deep PyTorch experience, strong systems debugging skills, and a security‑aware mindset.
Bonus Built scalable failure‑analysis and model‑transparency tooling (evaluation harnesses, telemetry analytics, interpretability workflows).
Background optimizing ML for real‑time or low‑latency environments.
Apply
Location:
San Francisco

We found some similar jobs based on your search