We are looking for an AI/ML Security Engineer with experience in prompt injection and AI security to join a remote team based in LATAM and working CST hours. You will help develop frameworks, tools, and infrastructure to evaluate, secure, monitor, and govern Machine Learning, Generative AI, and Agentic AI solutions. A key focus is identifying and remediating AI security threats, including prompt injection, jailbreaks, data leakage, adversarial inputs, and unsafe AI/agent behavior.
You will work closely with AI/ML, security, and product teams to build automated security evaluations, establish security metrics, investigate vulnerabilities, and implement effective remediation and guardrails across the AI/ML lifecycle.
Key Responsibilities
- Design and develop frameworks to evaluate the security, safety, accuracy, reliability, and performance of ML/GenAI solutions
- Identify, reproduce, and remediate prompt injection and indirect prompt injection vulnerabilities
- Develop automated security and adversarial tests for LLMs and Agentic AI, including jailbreaks, data leakage, and unsafe tool use
- Build and maintain metrics, monitoring, and evaluation capabilities for AI/ML applications
- Collaborate with engineering and security teams to develop AI threat models, security controls, and guardrails
- Investigate vulnerabilities and unexpected behavior across LLMs, AI agents, APIs, and supporting infrastructure
- Contribute to AI/ML infrastructure, CI/CD pipelines, APIs, and automation
- Write scalable, secure, and maintainable code, primarily in Python
- Develop unit, integration, functional, and security tests
- Participate in code reviews, troubleshooting, production support, and security assessments
- Stay current with emerging AI security threats, attack techniques, and defensive strategies
Requirements
- 4–5 years of experience in ML Engineering, Software Engineering, AI Security, Application Security, or a related field, or an advanced degree with relevant experience (preferred)
- Experience with Generative AI, LLMs, and/or Agentic AI (preferred)
- Understanding of LLM security threats, especially prompt injection, indirect prompt injection, jailbreaks, and data leakage (preferred)
- Experience with AI/ML security testing, adversarial testing, red teaming, vulnerability assessment, or threat modeling (preferred)
- Experience building AI/ML evaluation, monitoring, or governance frameworks (preferred)
- Strong Python programming skills; Java or other languages are a plus (preferred)
- Experience with ML deployment, CI/CD, data pipelines, and production infrastructure (preferred)
- Hands-on experience with AWS, Docker, and Kubernetes (preferred)
- Strong foundation in software engineering, APIs, distributed systems, databases, and testing (preferred)
- Experience with Git/GitHub and Agile/Sprint development environments (preferred)
- Strong debugging, analytical, and problem-solving skills (preferred)
Benefits
- 4–5 years of experience in ML Engineering, Software Engineering, AI Security, Application Security, or a related field, or an advanced degree with relevant experience (preferred)
- Experience with Generative AI, LLMs, and/or Agentic AI (preferred)
- Understanding of LLM security threats, especially prompt injection, indirect prompt injection, jailbreaks, and data leakage (preferred)
- Experience with AI/ML security testing, adversarial testing, red teaming, vulnerability assessment, or threat modeling (preferred)
- Experience building AI/ML evaluation, monitoring, or governance frameworks (preferred)
- Strong Python programming skills; Java or other languages are a plus (preferred)
- Experience with ML deployment, CI/CD, data pipelines, and production infrastructure (preferred)
- Hands-on experience with AWS, Docker, and Kubernetes (preferred)
- Strong foundation in software engineering, APIs, distributed systems, databases, and testing (preferred)
- Experience with Git/GitHub and Agile/Sprint development environments (preferred)
- Strong debugging, analytical, and problem-solving skills (preferred)
Originally posted on