About the role
This remote Senior Application Security Engineer engagement is open in Nigeria and forms part of a GenAI application-security evaluation and AI-training project. Its purpose is to assess how well static-analysis products detect application weaknesses in LLM agents and Retrieval-Augmented Generation (RAG) systems. Work uses Python, TypeScript, SAST, CodeQL, Semgrep, LLM agents, RAG and MCP. The engineer will also create small executable repositories illustrating Sensitive Information Disclosure (SID), Excessive Agency (EA), remediated implementations and near-miss cases.
What you will do
- Build working agent and RAG repositories whose reachable code contains Sensitive Information Disclosure and Excessive Agency weaknesses.
- Develop lifelike security examples using tool calling, agent memory, Model Context Protocol (MCP) and additional LLM-application elements.
- Create vulnerable, repaired and hard-negative variants while limiting changes to security-significant details.
- Assess static application security testing (SAST) tools for accurate detection of real vulnerabilities and avoidance of false alarms.
- Conduct source-to-sink analysis, following relevant data movement, action paths and trust boundaries.
- Document each security issue, including impacted assets, cause, safeguards, severity and risk that remains.
- Set out authorization contexts for actors, tenants, identities, authorized data, actions, purposes and recipients.
- Create deterministic tests that verify insecure, fixed and negative outcomes.
- Advise on relevant security controls and remediation options.
- Take part in technical calibration, peer assessment and quality-assurance work.
- Review code routes, uncover security problems and communicate the findings.
Requirements
- 5+ years of professional experience across application security, product security or security-centered software engineering.
- Hands-on experience developing LLM agents or RAG applications is required; the role also calls for experience building or protecting these systems.
- Substantial experience reviewing code securely and finding application-layer vulnerabilities.
- Practical command of taint analysis, source-to-sink analysis and static application security testing (SAST).
- Direct experience with analysis tools such as Semgrep or CodeQL.
- Knowledge of LangChain, LlamaIndex, Model Context Protocol (MCP), plus OpenAI or Anthropic SDKs.
- Strong knowledge of authorization concepts involving actors, identities, trust boundaries, tenant separation, access control, OAuth, IAM and identity management.
- Understanding of permitted information access and actions, purpose and recipient limitations, and controls at document level.
- Experience writing production software in Python and/or TypeScript.
- Skill in differentiating actual Sensitive Information Disclosure and Excessive Agency vulnerabilities from non-issues and false positives.
- Strong analytical capabilities and an ability to explain technical security results clearly.
How to apply
Apply using the employer’s official vacancy. The posting specifies no further application steps.