Senior Applied Scientist – Cyber Defense

This is a fully remote position, open to applicants in California.

📋 Description

• Collaborate with security experts to pinpoint high-impact workflows and spearhead the implementation of agentic systems for detection, investigation, and response.

• Provide technical guidance for intricate agentic AI projects across various teams.

• Develop context-aware agents that assess security data and institutional knowledge, utilize approved tools, and promote increased autonomy.

• Create a consistent model and agent evaluation framework using realistic security environments, curated datasets, analyst ground truth, and task-specific benchmarks.

• Assess end-to-end behavior through automated scoring, trajectory analysis, and adversarial testing.

• Enhance agent quality, reliability, and efficiency by leveraging evaluation results, production traces, and analyst feedback.

• Optimize models, retrieval processes, context, orchestration, and inference against measurable security outcomes.

• Transition AI capabilities from experimentation to production employing software engineering and MLOps/LLMOps methodologies.

• Integrate continuous evaluation, observability, versioning, controlled deployment, and safe rollback into the development lifecycle.

• Assess NVIDIA, open-source, frontier, and partner AI capabilities using a flexible, multi-model strategy.

• Provide evidence-based recommendations for adoption, adaptation, development, integration, or co-development.

• Convert technical insights into recommendations that shape architecture, Applied AI priorities, and partner roadmaps.

• Transform validated approaches into reusable capabilities for NVIDIA’s AI security ecosystem.

⛳️ Requirements

• BS, MS, or PhD in Computer Science, Artificial Intelligence, Machine Learning, Software Engineering, Cybersecurity, or a related technical field, or equivalent experience.

• Over 8 years of relevant experience in developing and deploying AI, machine learning, or intelligent software systems.

• Technical ownership of complex production initiatives.

• Strong software engineering competencies, particularly in Python.

• Proficiency in TypeScript or C#.

• Capability to design reliable and scalable systems beyond mere prototypes or experimental notebooks.

• Practical experience with large language models, retrieval-augmented generation, agentic architectures, agent harnesses, or similar methodologies.

• Experience in crafting AI evaluations and benchmarks utilizing curated datasets, ground truth, task-specific metrics, automated evaluators, error analysis, and expert feedback.

• Background in navigating AI capabilities through experimentation, deployment, monitoring, optimization, and continuous improvement using MLOps or LLMOps practices.

• Proven technical leadership across complex, cross-functional projects.

• Ability to make independent judgments, influence architecture and technical direction, and navigate ambiguous problems to achieve measurable results.

• Strong understanding of cybersecurity or experience applying AI and software engineering to security operations, detection, incident response, threat research, or similar adversarial fields.

• Extensive experience with evaluation environments and benchmarks for agentic systems, trajectory-level evaluation, task verifiers, adversarial scenarios, and safety or reliability testing.

• Experience in designing and calibrating LLM-as-a-Judge or other model-based evaluators.

• Familiarity with agentic architectures, orchestration systems, retrieval and context pipelines, or multi-agent strategies.

• Knowledge of NVIDIA AI technologies such as NeMo Evaluator, NeMo Gym, NeMo Agent Toolkit, NVIDIA NIM, Triton Inference Server, RAPIDS, or CUDA.

• Recognized contributions through open-source projects, benchmarks, publications, patents, conference presentations, or similar contributions in AI/cybersecurity.

🏝️ Benefits

• Equity

• Benefits