SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Veeam is seeking a Staff AI Security Engineer to join the VDC (Veeam Data Cloud) Platform Security team in San Jose. This is a high-impact individual contributor role focused on establishing AI/LLM security as a discipline across the organization.
You will design and implement data-handling controls for VDC's internal AI tooling and product AI features, including sensitive data redaction from prompts, model context, logs, and outputs before they reach third-party model providers. You'll build and productionize an AI-enhanced vulnerability reduction capability that integrates into CI/CD pipelines, surfaces real defects with low false-positive rates, and recommends or applies remediation while maintaining developer trust.
Key responsibilities include publishing and evolving secure-LLM-use patterns for engineering teams (input filtering, output validation, data classification, threat modeling), threat-modeling internal AI tools for prompt injection, indirect attacks, model exfiltration, and agent-tool boundary weaknesses. You'll partner with the red team on findings and build fixes with tool owners. You'll also collaborate with Compliance to shape auditor-facing evidence for AI-assisted controls and contribute the AI-security lens to adjacent programs including vulnerability management, supply chain risk, and compliance evidence collection.
You'll set VDC's direction on emerging LLM security tools and AI-forward workflows, deciding what to adopt, skip, or build in-house. The role works with Azure OpenAI Service, Anthropic Claude, OpenAI APIs, Microsoft Presidio, Cycode, Wiz, GitHub Actions, Azure DevOps, Python, Go, PowerShell, TypeScript, and Microsoft Sentinel.
This role requires 10+ years across security and engineering with recent focus on AI/ML systems in production, including LLM deployments, model risk, and AI-driven security. You should have hands-on experience shipping security controls at scale, deep knowledge of LLM threat models and mitigations, and the ability to influence cross-functional teams without direct authority.
