Security Architect - AI/ML

San R&D Business Solutions LLC•New York, NY

1d•Onsite

About The Position

The Security Architect – AI/ML leads the design and implementation of security controls across the full AI/ML lifecycle, including data ingestion, model training, deployment, and operational monitoring. This role establishes secure-by-design standards for LLMs, GenAI tools, model pipelines, and agent-based systems. The architect will develop AI security architecture patterns, governance frameworks, and risk controls while guiding engineering and leadership teams on secure AI adoption.

Requirements

10+ years of experience in Security Architecture or related cybersecurity roles.
Strong hands-on experience securing AI/ML systems and cloud-native environments.
AI/ML lifecycle security
LLM and Generative AI risk mitigation
Threat modeling for AI systems
Data security and privacy controls
Prompt injection
Data poisoning
Model extraction
Adversarial ML attacks
Strong experience with cloud platforms (AWS, Azure, or GCP).
NIST 800-53
NIST AI RMF
FedRAMP
FIPS 199/200
MITRE ATLAS
Excellent communication and cross-functional leadership skills.

Nice To Haves

Experience integrating AI telemetry into SOC/SIEM platforms.
Background in DevSecOps and secure CI/CD pipeline architecture.
Experience with enterprise data classification and DLP strategies.
Industry security certifications (CISSP, CCSP, or equivalent).
Experience working in regulated industries (financial services, healthcare, government).

Responsibilities

Design and implement security controls across AI/ML systems lifecycle.
Develop secure architecture patterns for LLMs, GenAI, and agent-based systems.
Establish governance and risk management frameworks for AI solutions.
Evaluate third-party AI services for security and compliance risks.
Conduct AI-specific threat modeling (prompt injection, data poisoning, model extraction, adversarial attacks).
Lead AI red teaming exercises and security validation testing.
Implement guardrails, monitoring, and model behavior controls.
Define AI security baselines aligned with Zero Trust principles.
Secure MLOps and LLMOps pipelines including CI/CD controls.
Implement controls for model versioning, artifact protection, and access governance.
Integrate AI telemetry and monitoring into enterprise security operations (SOC/SIEM).
Ensure secure data pipelines and protection of training datasets.
Align AI systems with regulatory and privacy requirements.
Implement controls aligned with FedRAMP, NIST 800-53, NIST AI RMF, FIPS 199/200, and MITRE ATLAS.
Produce architecture documentation, security standards, and policies.
Provide strategic guidance to engineering and executive leadership.