Agentic AI Security Engineer, Lead

Job Number: R0252003

About the role

As organizations adopt AI systems that can reason, use tools, and take action, they need engineers who understand both how these systems work and how to secure them. We're looking for an AI/ML engineer who can combine hands-on development, security expertise, and architectural thinking to help our customers deploy agentic AI securely and responsibly.

As a security-focused AI/ML Engineer on our team, you'll design, build, and evaluate secure AI applications and agentic workflows. You'll examine how models, agent orchestration, tools, data sources, and enterprise infrastructure interact, identify security risks, and develop practical controls throughout the system lifecycle. You'll turn emerging research and complex technical requirements into working prototypes, reusable architecture patterns, and clear technical guidance. You'll collaborate with AI engineers, cybersecurity specialists, architects, and mission stakeholders to develop solutions that balance capability, security, and operational needs. Your work will help technology leaders and implementation teams understand how to move from AI experimentation to secure, governed deployment.

Responsibilities

• Design and validate secure architectures for generative AI applications, AI agents, and multi-agent systems, including model access, orchestration, tool execution, memory, and retrieval augmented generation
• Develop prototypes and reference implementations to demonstrate security controls and evaluate architectural tradeoffs
• Assess AI systems for risks such as prompt injection, sensitive data exposure, excessive permissions, unsafe tool execution, and compromised integrations
• Design and implement controls for agent identity, least privilege access, execution isolation, tool authorization, data protection, and human approval of consequential actions
• Develop architecture diagrams, data flow diagrams, trust boundary models, and technical documentation that explain system behavior, control placement, and implementation requirements
• Establish evaluation and monitoring approaches for agent behavior, security, reliability, and policy compliance, including traceability across model interactions and tool calls
• Translate security and governance requirements into engineering practices, including testing, version control, change management, audit trails, and rollback procedures
• Evaluate emerging models, agent frameworks, integration protocols, and security tools, and communicate findings and recommendations to technical and non-technical audiences
• Provide day-to-day technical leadership to engineers
• Conduct design reviews, guide architectural decisions, and raise the overall technical bar of the team

Requirements

• 8+ years of experience in software engineering or AI/ML engineering, including developing and deploying AI/ML systems
• Experience developing in Python and applying software engineering practices, including Git, automated testing, code review, and dependency management
• Experience building applications using large language models, including tool use, retrieval augmented generation, and agent orchestration
• Experience designing or evaluating system architectures involving APIs, data pipelines, model services, and enterprise integrations
• Experience applying cybersecurity principles to software or AI systems, including identity and access management, trust boundaries, secure data handling, and threat modeling
• Experience with traditional ML approaches and algorithms and their applications to cybersecurity
• Knowledge of NIST security and risk management frameworks, including SP 800-53 and AI Risk Management Framework, and Zero Trust Architecture principles, including identity-based access, least privilege, and continuous verification
• Knowledge of AI-specific security risks and approaches for evaluating and mitigating vulnerabilities in generative AI and agentic systems
• Ability to obtain a Secret clearance
• Bachelor's degree in CS, Mathematics, Engineering, or Cybersecurity

Preferred Qualifications

• Experience developing agentic AI applications using frameworks or platforms such as Strands Agents, LangGraph, Amazon Bedrock, GraphRAG, or comparable technologies, including integrating agents with workflow orchestration tools such as Temporal
• Experience securing tool integrations using Model Context Protocol or other agent communication and integration mechanisms
• Experience conducting AI red teaming, adversarial testing, or automated security evaluations of models and agentic workflows
• Experience implementing AI gateways, policy enforcement, execution sandboxes, agent observability, or human-in-the-loop controls
• Experience deploying AI applications in cloud or on-premises environments using containers, CI/CD pipelines, and Infrastructure-as-Code
• Experience with model training, fine-tuning, evaluation, or serving using frameworks such as PyTorch or comparable technologies
• Experience with distributed computing, data engineering, feature pipelines, or relational and vector databases
• Experience applying NIST guidance and Zero Trust Architecture principles to AI systems, and translating security requirements into architectural controls, implementation patterns, and assessment evidence
• TS/SCI clearance

Clearance

Applicants selected will be subject to a security investigation and may need to meet eligibility requirements for access to classified information.

Pay

The projected compensation range for this position is $99,000.00 to $225,000.00 (annualized USD). The estimate displayed represents the typical salary range for this position and is just one component of Booz Allen's total compensation package for employees. Salary is determined by various factors, including but not limited to location, the individual's particular combination of education, knowledge, skills, competencies, and experience, as well as contract-specific affordability and organizational requirements.

Benefits

• Health, life, disability, financial, and retirement benefits
• Paid leave
• Professional development
• Tuition assistance
• Work-life programs
• Dependent care
• Recognition awards program for exceptional performance and superior demonstration of values

Full-time and part-time employees working at least 20 hours a week on a regular basis are eligible to participate in Booz Allen's benefit programs. Individuals that do not meet the threshold are only eligible for select offerings, not inclusive of health benefits.