$ cat jobs/ai-security-in-engineering-teaching-expert-b2b-part-time-tripleten-2e5505e323cb.json
AI Security in Engineering Teaching Expert (B2B, Part-time)
Description Nebius Academy (powered by TripleTen ) partners with forward-thinking companies to help technical teams adopt AI safely and build practical, job-ready skills that drive measurable business impact. We design hands-on learning experiences that help engineers apply AI tools in real-world workflows while understanding the security risks introduced by agentic systems, tool use, delegated permissions, data access, and AI-powered development environments. As we continue expanding our AI Security portfolio, we're looking for an experienced Instructor AI Security in Engineering to deliver highly practical workshops for engineering teams. This role combines deep hands-on expertise in AI and agentic security with technical facilitation, helping experienced engineers understand and mitigate risks related to prompt injection, MCP trust boundaries, permissions, AI supply-chain vulnerabilities, secure AI-generated code, and data governance. What you will do Deliver hands-on AI Security workshops for experienced engineering teams, combining concise concept explanations with practical labs on real repositories. Demonstrate real-world attack scenarios and mitigations involving agentic AI coding tools, including indirect prompt injection, delegated permissions, autonomous command execution, and workspace trust. Guide learners through MCP security practices, including trust boundaries, read/write scoping, least-privilege configurations, and re-approval of changed tool definitions. Help engineers secure AI-generated code and pull requests using DevSecOps controls such as secret scanning, SAST, SCA, branch protection, CODEOWNERS, and repository-level security instructions. Explain and demonstrate AI-specific supply-chain risks, including hallucinated dependencies, poisoned tool descriptions, and compromised MCP servers. Facilitate practical discussions around AI data governance, including sensitive data exposure, credentials, proprietary code, model data flows, retention, and