Member of Technical Staff - Imagine Safety
Core
Building critical safety systems and infrastructure to ensure Grok's multimodal generation capabilities (images, video, audio) are responsibly deployed while preserving creativity.
Role type
Product Safety Engineer (Multimodal AI)
Builds
Scalable safety systems, detection systems, evaluation frameworks, and feedback loops for Grok's media generation platform.
Domain
Generative AI, Multimodal Media, Content Safety
Deliverable
production ML models | product features | infrastructure
Required skills
Rust, production safety/trust & safety systems, real-time detection systems, data pipelines, evaluation frameworks, high-throughput AI applications, robust solution delivery, complex problem solving
Preferred skills
multimodal content safety, generative AI safety, machine learning classifiers, safety evaluation, red-teaming, adversarial testing, distributed systems, real-time inference serving, Kubernetes, observability tools, large-scale data infrastructure, cross-functional collaboration, content moderation, anti-abuse, model alignment
Technologies
Rust, Kubernetes, observability tools
Responsibilities
Design and implement scalable safety systems for real-time content moderation and risk detection; Build infrastructure to measure, monitor, and mitigate safety risks like harmful content, bias, deepfakes, and IP issues; Develop tools and pipelines for rapid iteration on safety policies; Architect feedback loops between user interactions, model outputs, and training data; Own full-cycle development of safety features from prototyping to deployment and monitoring; Partner with engineering, research, and product teams to embed safety into core features.
Seniority
Mid-to-Senior, hands-on IC