Senior AI Engineer, AI Safety

NVIDIA•Santa Clara, CA

3d•$184,000 - $356,500

About The Position

The Senior AI Engineer will join NVIDIA’s AI Safety Engineering team to scale methods for securing and aligning advanced AI systems. The role focuses on architecting and developing safety infrastructure, especially autonomous automation workflows, in partnership with Research and Model Development teams. Work includes developing models for harm detection, content moderation, dynamic guardrails, post-training, and automated research pipelines. The engineer will build agentic automation for safety evaluations, red-teaming, model testing, post-training, and ablations; develop specialized safety systems; and create automated research frameworks using LLMs and autonomous agents to find vulnerabilities, generate or augment safety datasets, and propose alignment techniques. The role also involves turning research concepts and proof-of-concept interventions into robust, scalable production solutions, while collaborating with researchers, platform engineers, and data scientists.

Requirements

Bachelor’s or master’s degree in Computer Science, Machine Learning, Software Engineering, or a related field, with equivalent experience.
At least 8 years of experience in ML engineering or software engineering.
Demonstrated experience building complex automated workflows using LLMs, including multi-turn tool-calling, orchestration, and autonomous agent loops.
Deep proficiency in one or more programming languages such as Rust or Python.
Experience with machine learning frameworks such as PyTorch.
Ability to write scalable, maintainable, and highly optimized code.
Demonstrated ability to tackle open-ended, ambiguous problems through experimentation, iteration, and persistence.
Ability to take projects from an initial research idea through deployment of a scalable production system.

Nice To Haves

Prior experience in Trust & Safety, cybersecurity, threat intelligence, adversarial machine learning, or content moderation.
Hands-on experience developing, fine-tuning, or deploying models for harm detection, content safety filtering, or run-time AI guardrails.
Experience with filtering content involving violence, hate, harassment, or CBRNE.
Contributions to projects using LLMs to automate coding, data synthesis, or scientific research.
Active contributions to publicly available agent frameworks such as LangChain, AutoGPT, or Semantic Kernel, or to AI safety benchmarking tools.

Responsibilities

Compose, build, and deploy agentic automation loops for safety evaluations, red-teaming, model testing, model post-training, and ablations.
Train and implement specialized systems for identifying harmful content, coordinating content quality, and providing run-time guardrails.
Build and guide automated research frameworks that use LLMs and autonomous agents to discover vulnerabilities, generate or augment safety datasets, and propose alignment techniques.
Engineer proof-of-concept safety interventions and research ideas into robust, scalable solutions integrated with production pipelines.
Partner with researchers, platform engineers, and data scientists to translate AI safety research into practical automated systems.
Own projects from initial research whiteboarding through deployment of scalable production systems.