AI/LLM Safety Engineer

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

📋 Description

• Take responsibility for the safety of models and agents during production, concentrating on AI Safety, Trust & Safety, and Responsible AI.

• Create and uphold adversarial prompt sets, scenario-based testing suites, and regression suites.

• Ensure validation of every model and agent update prior to release.

• Facilitate red-teaming exercises addressing jailbreaks, prompt injection, tool misuse, and data exfiltration.

• Document observations and guide issues through remediation and resolution.

• Develop and refine input/output filtering, tool-boundary constraints, action validation, sensitive-data redaction, and policy prompting.

• Incorporate safety checks into CI/CD pipelines and runtime systems.

• Threat-model agentic scenarios that involve tool-call boundaries, sandbox isolation, and least-privilege access principles.

• Avert data exfiltration and irreversible actions via chained tool calls.

• Evaluate reinforcement-learning environments and trajectory data alongside environment and agent engineering teams.

• Enhance AI features with structured logging, tracing, and metrics.

• Compile test reports, evaluation summaries, and mitigation validation for governance reviews.

• Work collaboratively with Product, UX, Research, and Data teams on safety interactions and evaluation objectives.

⛳️ Requirements

• A Bachelor's or Master's degree in Computer Science, Software Engineering, Cybersecurity, or a related technical field—or equivalent practical experience.

• Over 4 years of experience in building production software, with direct involvement in securing ML/LLM systems.

• Proficient software engineering capabilities with the ability to write production-quality code, primarily in Python.

• Comprehensive understanding of LLMs and ML, covering model operation, prompt engineering, fine-tuning safety implications, and RAG risks.

• Security-focused mindset with proven threat-modeling skills.

• Knowledge of access control, data retention, and incident response fundamentals.

• Awareness of prompt injection, jailbreaks, data poisoning, supply-chain risks, and OWASP LLM Top 10.

• Practical experience with safety evaluations or red teaming, including presenting a genuine finding and its remediation.

• Preferred: experience with garak, PyRIT, promptfoo, Giskard, and NeMo Guardrails.

• Preferred: publications, open-source contributions, or responsible disclosures in AI safety or security.

• Preferred: familiarity with NIST AI RMF, ISO/IEC 42001, and EU AI Act.

• Preferred: engineering experience with agents, RL environments, and/or tool usage.

• Preferred: experience with MITRE ATLAS and STRIDE/PASTA.

🏝️ Benefits

• Opportunities for learning, development, and capability expansion.

• Significant prospects for career advancement.

• Chances to pursue leadership roles and broaden career paths.

• Support for professional growth and innovation.

• Opportunity to make a significant impact.

• Flexible, empowered decision-making and ownership of initiatives.