Job Summary
This position involves the creation of functional repositories that intentionally incorporate security vulnerabilities within LLM agents and RAG systems to assess static analysis tools. The role also encompasses conducting source-to-sink analysis, establishing authorization contexts, and documenting security findings to enhance the reliability of AI evaluations.
Responsibilities
- Develop runnable repositories with genuine security vulnerabilities within LLM agents and RAG systems.
- Evaluate and assess static analysis tools for their effectiveness in identifying vulnerabilities.
- Conduct comprehensive source-to-sink analysis.
- Define and outline authorization contexts relevant to security evaluations.
- Document and report security findings to contribute to the reliability of AI evaluation processes.
Qualifications
- Minimum of 5 years of experience in application security or security-driven software engineering.
- Hands-on experience with programming languages, specifically Python or TypeScript.
- Expertise in security analysis tools such as CodeQL or Semgrep.
- Prior experience in securing LLM agents or RAG applications is essential.
