deepjobsPost a job
← All jobs

Research Scientist, PAR Safety

Meta

Where
Menlo Park, CA
Salary
$154k–217k
Apply

Meta is seeking a Research Scientist to advance the state of the art in artificial intelligence, contributing to foundational and applied research that powers Meta's AI products and platforms. In this role, you will design and develop novel machine learning algorithms, models, and systems across areas such as large language models, multimodal learning, reinforcement learning, computer vision, or related AI domains. Your research will directly influence the capabilities of Meta AI and the broader AI infrastructure that serves billions of people worldwide.

Responsibilities

Design and conduct original AI research, developing novel algorithms, model architectures, and training methodologies that advance the state of the art in machine learning

Translate research findings into scalable, production-ready systems by collaborating with engineering and product teams to integrate new methods into Meta's AI platforms

Define and drive the technical direction for a significant research area, including setting research goals, identifying key open problems, and building a roadmap to address them

Develop rigorous experimental frameworks including evaluation benchmarks, ablation studies, and statistical analyses to validate research hypotheses and measure model performance

Architect reusable research infrastructure, training pipelines, and evaluation tooling that accelerates the productivity of the broader research organization

Identify and instrument the right metrics and telemetry to monitor model behavior, detect regressions, and guide iterative improvements across research and production systems

Collaborate with cross-functional partners including product managers, engineers, and data scientists to translate product requirements into well-scoped research problems with measurable outcomes

Communicate research findings through technical documentation, internal presentations, and external publications such as peer-reviewed conference papers and journal articles

Mentor other researchers and engineers on the team, providing technical guidance, constructive feedback, and frameworks to support their growth and the quality of shared research outputs

Proactively identify opportunities to improve research processes, tooling, and team workflows, and drive changes that increase the efficiency and impact of the research organization

Minimum qualifications

Bachelor's degree in Computer Science, Computer Engineering, relevant technical field, or equivalent practical experience

Advanced degree (PhD) in Computer Science, Machine Learning, Statistics, or a closely related technical field, or equivalent depth of research experience

2+ years of experience conducting research in machine learning, deep learning, or a closely related AI discipline, with demonstrated contributions to original or materially improved methods

Proven track record of delivering research with measurable impact, evidenced by peer-reviewed publications, adopted open-source methods, or deployed AI systems at scale

Hands-on experience developing, training, debugging, and evaluating machine learning models using Python and deep learning frameworks such as PyTorch or JAX

Experience independently defining technical direction and leading complex, multi-stakeholder research initiatives from problem formulation through evaluation and deployment

Experience communicating complex technical concepts and influencing research or engineering decisions through written proposals, design documents, and cross-functional collaboration

Preferred qualifications

Experience scaling machine learning research to production systems, including distributed training, efficient inference, or model optimization under latency and cost constraints

Expertise in one or more specialized AI domains such as large language models, multimodal learning, reinforcement learning, computer vision, or generative modeling

Demonstrated ability to build and improve research infrastructure, evaluation benchmarks, or synthetic data pipelines that accelerate team-wide research productivity

Listed on www.metacareers.com