Job Summary
The Lead Software / MLSecOps - AI Security Platform is responsible for designing, implementing, and maintaining secure AI/ML pipelines and strengthening the security posture of internally developed products. This role focuses on securing AI models, coding agents, autonomous workflows, and AI-assisted software development platforms throughout the software development lifecycle. The position combines software engineering, ML security, DevSecOps, threat assessment, and AI governance to identify and mitigate emerging security risks while enabling development teams to adopt secure AI practices.
Key Responsibilities
• Design, implement, and maintain secure ML pipelines for AI/ML model evaluation, validation, deployment, and inference.
• Assess and mitigate security risks throughout the ML lifecycle, including data ingestion, model storage, and deployment.
• Evaluate, secure, and govern AI coding agents, autonomous agents, and agentic workflows used throughout the software development lifecycle.
• Define and implement security guardrails for AI-assisted software development platforms, agent orchestration frameworks, and autonomous development pipelines.
• Develop and operationalize an Agentic Development Lifecycle (ADLC) framework incorporating security requirements, threat modeling, testing, deployment governance, and continuous security monitoring.
• Evaluate AI Security Posture Management (AI-SPM) capabilities and establish processes to identify, categorize, prioritize, and remediate risks associated with AI assets, agents, prompts, datasets, and AI-enabled applications.
• Assess emerging threats affecting frontier AI models and agentic systems and recommend preventive and detective security controls to reduce enterprise risk.
• Develop and maintain AI/ML pipeline code using Python and CI/CD, ensuring robust security controls and adherence to best practices.
• Integrate security scanning of AI/ML models into development workflows using a shift-left approach, interpret scan results, and remediate identified issues.
• Evaluate and optimize model inference deployment strategies while balancing security, performance, and resource utilization.
• Monitor vulnerabilities affecting machine learning models, Large Language Models (LLMs), and AI agents, including prompt injection, data poisoning, model theft, and adversarial attacks.
• Collaborate with data scientists, ML engineers, and security teams to promote secure ML practices.
• Establish strong partnerships with stakeholders across technology and product organizations.
• Perform other duties as required.
Required Qualifications
• Bachelor's degree in Computer Science, Information/Cyber Security, Computer Systems Engineering, Computer Information Systems, or equivalent education and experience.
• 8+ years of experience in IT or cybersecurity roles, including 5+ years of experience specifically in software engineering roles.
• Hands-on experience with MLOps pipelines and model deployment tools such as Kubeflow, MLflow, or SageMaker.
• Strong programming skills in Python and experience with CI/CD for automation and pipeline development.
• Hands-on experience with AI coding assistants and coding agents such as GitHub Copilot, Microsoft Copilot, Cursor, Claude Code, Windsurf, or similar platforms.
• Experience using AI-driven development techniques across programming languages such as Python, Java, JavaScript, C#, .NET, and Go.
• Strong understanding of agentic architectures, AI agents, autonomous workflows, Retrieval-Augmented Generation (RAG), and associated security considerations.
• Experience assessing agent sandboxes, agent runtime environments, agent-to-tool communications, and agent execution workflows.
• Understanding of security controls governing agent behavior, including permissions, approval workflows, runtime controls, and data protection guardrails.
• Ability to identify, categorize, prioritize, and operationalize risks associated with AI assets, models, prompts, datasets, agents, and AI-enabled applications.
• Strong understanding of Agentic Development Lifecycle (ADLC) principles and the integration of security controls across planning, development, testing, deployment, and operations.
• Knowledge of frontier AI model cybersecurity programs and layered controls for mitigating AI-driven cyberattacks, adversarial ML threats, and model and agent abuse.
• Familiarity with structured data systems, including SQL and data warehouses, and unstructured data systems, including object storage and NoSQL databases.
• Familiarity with Databricks.
• Experience with ML security tools for model scanning and vulnerability assessment.
• Knowledge of OWASP AI/ML vulnerabilities, including prompt injection, data and model poisoning, model extraction and inversion, adversarial example attacks, and ML component supply chain risks.
• Understanding of AI/ML model formats such as pickle, TensorFlow, and safetensors.
• Experience implementing model scanning solutions as part of model development.
• Understanding of CI/CD pipelines covering source control, integration, and deployment, using tools such as Bitbucket, Jenkins, Jira, Artifactory, Nexus, Git, and SonarQube or Snyk scanners.
• Strong understanding of internet technologies and web development best practices.
• Understanding of advanced iterative Agile methodologies, cloud and container security, and Generative AI security.
• Strong analytical, problem-solving, and cross-functional collaboration skills across development and security disciplines.
• Ability to communicate cybersecurity and AI/ML security controls effectively to technical and non-technical stakeholders.
• Excellent communication and presentation skills.
• Ability to work independently, collaborate effectively with remote teams, and maintain a positive, self-motivated approach.
Preferred Qualifications
• Previous software engineering or architecture experience with Java, C#, .NET, JavaScript, or Python.
• Experience developing RESTful web services.
• Familiarity with microservices architecture and design patterns.
• Experience with cloud security, container security, and advanced iterative Agile practices.
• Certifications such as CEH, CISSP, CSSLP, GCIA, GPEN, or GWAPT.
