Sr Offensive AI Security Engineer II

About the Role

Samsara is building a new role at the intersection of security engineering and applied AI. You'll be the most AI-forward technologist on our Application Security team, using frontier AI models to find vulnerabilities across cloud, software, and firmware before attackers do. You'll also evaluate each new model for security use and help build guardrails for adopting AI safely.

*Note: This role is open to candidates residing in the US except the San Francisco Bay Metro Area, NYC Metro Area, and Washington, D.C. Metro Area.*

Responsibilities

• Deliver accurate, consistent security assessments of Samsara's software, code, and firmware, evaluating resilience against AI-driven/adversarial attacks, and implementing fixes.
• Create and translate general AI security findings into Samsara-specific fixes for our application security layer, adapting techniques built for software to hardware and embedded firmware.
• Find both AI-specific and traditional vulnerabilities (e.g., injection, IDOR) using AI-driven offensive validation.
• Build and pressure-test novel AI-enabled security tooling and workflows, driving adoption across the broader security organization.
• Benchmark each new frontier model against the previous one for security tasks, establishing release criteria and guardrails.
• Support AppSec vulnerability triage through an office-hours rotation, and help run the bug bounty program.
• Partner cross-functionally with engineering, hardware, and leadership, presenting findings to security and executive leadership.
• Help define and drive Samsara's medium- and long-term strategy for adopting AI safely and effectively within security operations.
• Use frontier AI/LLM tools hands-on for both offensive validation (red-teaming) and defensive tooling.

Minimum Requirements

• 8+ years of experience in application security, vulnerability management, threat modeling, offensive security, or a closely related discipline, with demonstrated application fix impact.
• Hands-on experience building AI or agentic workflows for security use cases.
• Experience conducting technical security assessments (e.g., offensive security or validation exercises) across both software and hardware/firmware environments.
• Experience in creating solutions for secure code with engineers using languages such as Go.
• Experience developing AI and agentic solutions using GitHub and developing within AWS cloud.
• Required experience integrating, securing, or red-teaming AI-enabled systems (e.g., LLMs, agentic workflows, prompt injection/adversarial testing).

Preferred Qualifications

• Experience with AI-specific attack techniques (prompt injection, jailbreaks, model/agent abuse) for vulnerability management and application security.
• Experience building or validating security controls for agentic AI, RAG systems, or LLM-integrated platforms.
• Experience red-teaming LLM-enabled systems.
• Experience evaluating or benchmarking LLMs.
• Bug bounty platform experience (e.g., Bugcrowd).
• Experience in federal or FedRAMP-style environments.
• Track record of driving org-wide security strategy or presenting technical findings to senior/executive stakeholders.
• Relevant industry certifications (e.g., OSCP, GIAC, CISSP) or AWS certifications.