Что будет проверяться
Lead security for Replit's AI coding products, pioneering research on the security of AI-generated ("vibe coded") software and working directly with the Agent to make its code generation safer for millions of developers. The role blends traditional application and cloud security with novel LLM-specific threat modeling like prompt injection and insecure code synthesis. A technical interview would probe web and cloud vulnerability classes, securing multi-tenant code-execution sandboxes, and how to detect and mitigate security flaws introduced by an LLM at generation time.
ExoForm не аффилирован с Replit. Это независимая тренировочная страница.