About the Role
We are seeking skilled GenAI Application Security Engineers to contribute to the development and assessment of security scenarios involving AI-powered applications. In this role, you will work with real-world GenAI systems to identify security weaknesses, reproduce vulnerabilities, implement effective fixes, and validate the effectiveness of security controls.
You will be responsible for analyzing application architectures, investigating potential attack paths, creating proof-of-concept scenarios, and producing clear technical documentation supported by reproducible tests.
Key Responsibilities:
- Identify, develop, or obtain GenAI-based applications, including systems built around AI agents, Retrieval-Augmented Generation (RAG), and external tool integrations.
- Introduce controlled security weaknesses into application environments and develop corresponding secure implementations.
- Investigate vulnerabilities by following data flows, execution paths, callbacks, background processes, and component interactions across multiple files.
- Produce detailed security reports covering affected code sections, vulnerability conditions, security control limitations, exploitation scenarios, and recommended remediation strategies.
- Define and assess access control requirements, considering user identities, organizational boundaries, tenant separation, data permissions, intended purposes, and authorized actions.
- Develop targeted security tests and proof-of-concept exploits to reproduce vulnerabilities and verify that proposed fixes successfully address them.
- Evaluate how AI models interact with application components and external tools, using live model execution whenever technically feasible.
- Analyze security mechanisms and identify situations where existing protections fail to prevent unauthorized access or unintended actions.
Required Qualifications:
- At least 3 years of professional experience in application security, product security, penetration testing, secure code auditing, or vulnerability analysis.
- Strong ability to investigate complex application behavior and follow data from its origin to potentially vulnerable execution points across different components.
- Solid understanding of identity management, access control, privilege management, tenant isolation, security boundaries, and approval mechanisms.
- Familiarity with the OWASP Top 10 for LLM Applications, particularly LLM02 (Sensitive Information Disclosure) and LLM03 (Excessive Agency), as well as prompt injection techniques and improper tool execution.
- Knowledge of common vulnerability classifications and their corresponding CWE identifiers, including CWE-200/201, CWE-532, CWE-284/285/862/863, and CWE-250/269.
- Practical experience creating security validation scenarios, including exploit reproduction, regression testing, and comparisons between vulnerable and remediated implementations.
- Experience with static application security testing tools such as CodeQL and Semgrep, including the ability to interpret SARIF reports and recognize vulnerabilities that automated pattern matching may overlook.
- Hands-on familiarity with GenAI development and security ecosystems, including LLM function calling, autonomous agents, MCP, RAG pipelines, memory systems, and observability or tracing solutions.
- Experience with frameworks and technologies such as LangChain, LlamaIndex, OpenAI SDK, Anthropic SDK, and OpenTelemetry.
- Strong software engineering fundamentals, including production-level coding, Docker, Git, source code investigation, and reviewing security-related commits and upstream patches.
- Excellent technical communication skills and the ability to produce clear documentation in English.
Offer Details:
- Remote
- Full-time dedication: 40 hours week with 4 hours PST overlap
- Location: Bangladesh, Brazil, Colombia, Egypt, Ghana, India, Indonesia, Turkey and USA
- Engagement Length: 4 weeks
Please note: This position is paid on a task based model. The hourly rate listed in this form is only an estimated average to provide compensation context. The final payment will depend on the tasks completed.
Important note: Due to the high volume of applications, we are unable to provide individual feedback at this stage. Only candidates shortlisted for the next steps will be contacted by our team.