Senior Lead Security Architect, AI/ML Platforms
On-sitePlano, Texas, United States
Job Summary
Develop and enhance security strategies, red teaming programs, and solution designs for AI applications, agents, and agentic workflows. Design secure architectures by reviewing code to ensure adversarial resilience and secure-by-default patterns. Reduce vulnerabilities in LLMs and agents through threat modeling, adversarial testing, and evolving policies across the model development lifecycle. Collaborate with product, engineering, and risk stakeholders to identify emerging threats and implement scalable controls. Define agent security controls including authentication, authorization, and governance for tool registration. Provide guidance on logging, monitoring, and observability for AI applications to support detection and forensics. Evaluate orchestration patterns to ensure safe execution boundaries and reliable policy enforcement. Assess integration patterns for Model Context Protocol and similar mechanisms to prevent data exfiltration. Work with cloud teams to secure infrastructure configuration and alignment with enterprise security architecture. Engage with external researchers to translate best practices into engineering guardrails. This role is designated as a High Risk Role (HRR) and requires additional pre-hire screening.
Required Qualifications
- 5 years of applied experience in cybersecurity architecture and/or securing AI/ML systems, including architecture reviews and risk-based control design.
- Practical cloud-native experience in AWS, GCP and/or Azure, with hands-on experience using Public Cloud AI/ML services (e.g., SageMaker, Bedrock) and applying enterprise security patterns in production environments.
- Advanced proficiency in one or more programming languages or applications, with the ability to review code and architecture for security and resilience concerns.
- Advanced knowledge of cybersecurity architecture, applications, and technical processes, with considerable in-depth knowledge in artificial intelligence and machine learning.
- Experience with AI and machine learning concepts and technologies, including notebooks, Python, TensorFlow, PyTorch, and common ML development workflows.
- Solid understanding and practical experience across the model development lifecycle (MDLC), including data acquisition and preparation, model experimentation, training and testing, serving, and MLOps.
- Solid understanding of the AI system attack surface, threats, and mitigating controls across the MDLC, including AI-specific risks such as prompt injection, training data compromise, unsafe output handling, and retrieval risks.
- Working knowledge of AI agent security/safety fundamentals, including authN/authZ, secure tool/skill use, least-privilege execution, secure context and memory handling, and requirements for logging and observability suitable for audit and incident response.
- Knowledge of AI safety, AI alignment, and AI cybersecurity concepts and trends, with the ability to translate evolving threats into practical engineering controls.
- This role is designated as a High Risk Role (HRR) and is subject to additional pre-hire screening and/or role-based requirements in accordance with applicable firm policies
Desired Qualifications
- Practical experience designing, developing, or securing AI agents following security best practices, including safe orchestration patterns, secure tool connectivity, and controlled autonomy.
- Experience with API security + IAM/enterprise authorization, including authentication, authorization, abuse-prevention controls for AI-facing and agent-facing APIs, and OAuth 2.0, OpenID Connect, and SAML.
- Knowledge of containers and container orchestration (Docker, Kubernetes, Helm) and the security implications of runtime isolation and workload identity.
- Knowledge of cloud infrastructure as code (IaC) (Terraform), including secure-by-default patterns and control enforcement.
- Knowledge of networking concepts and protocols (TCP/IP, routing, DNS, DHCP) and how these affect secure deployment and segmentation of AI systems.
- Familiarity with MCP and/or other agent tool-connection standards, including security implications of tool discovery, trust boundaries, and authorization delegation.
- Preferred certifications (one or more): AWS Certified Machine Learning – Specialty, Microsoft Certified: Azure Data Scientist Associate, AWS Certified Security – Specialty, Microsoft Certified: Cybersecurity Architect Expert, and/or CISSP.
Hiring someone like this?
Get your role in front of qualified candidates on Sorce.