Principal Agentic AI Security Engineer

Hewlett Packard Enterprise • Durham, NC

2d • $126,500 - $290,500 • Hybrid

About The Position

The Agentic AI Engineer designs, builds, and operates production-grade agentic systems that combine foundation models, software services, enterprise data, tools, and automation. The role focuses on system architecture, agent orchestration, model-agnostic AI integration, deterministic execution, and engineering secure and reliable platforms for complex multi-step workflows. Systems should treat models, frameworks, and inference platforms as replaceable components, so solutions can use commercial, cloud-hosted, or open-weight models according to workload requirements.

The position is hybrid, requiring an average of two days per week at an HPE office. Successful candidates should have deep expertise in at least two of five technical concentration areas—system design and architecture; agentic systems and harness engineering; model-agnostic AI and context engineering; platform engineering and production reliability; and evaluation, security, and AI safety—and strong working knowledge across the remaining areas.

Hewlett Packard Enterprise describes itself as a global edge-to-cloud company. It offers benefits intended to support physical, financial, and emotional wellbeing, as well as programs for personal and professional development. HPE also emphasizes an inclusive and accessible workplace.

Requirements

• Bachelor's or Master's degree in Computer Science, Engineering, Artificial Intelligence, or a related technical discipline, or equivalent practical experience.
• Typically 5–8 years of software engineering, platform engineering, distributed systems, AI engineering, or related experience.
• Demonstrated experience designing and delivering production software systems.
• Deep expertise in at least two of the five listed technical concentration areas and strong working knowledge across the remaining areas.
• Knowledge of distributed systems, microservices, APIs, event-driven and asynchronous architectures, workflow and state management, concurrency, scalability, fault tolerance, resiliency, and failure recovery.
• Knowledge of agent orchestration, multi-agent architectures, planning, tool calling, delegation, state, memory, long-running workflows, MCP, A2A, tool registries, skills, execution harnesses, sandboxed execution, and human-in-the-loop patterns.
• Knowledge of foundation model APIs and open-weight or self-hosted models, model abstraction and routing, fallback, benchmarking, workload-based model selection, structured generation, tool/function calling, RAG, hybrid retrieval, embeddings, re-ranking, vector databases, and context management.
• Strong Python and backend software engineering skills, including REST/gRPC APIs, asynchronous services, messaging, and distributed processing.
• Knowledge of Docker, Kubernetes, CI/CD, cloud infrastructure, logging, metrics, tracing, observability, production troubleshooting, performance, scalability, reliability, and resource optimization.
• Knowledge of agent and model evaluation, regression testing, task-success measurement, tool-use and trajectory evaluation, AI security, prompt injection, tool abuse, and data-leakage protections.
• Knowledge of least privilege, identity, authorization, sandboxing, policy enforcement, guardrails, human approval gates, monitoring, traceability, and auditability.

Nice To Haves

• Hands-on experience developing LLM-powered or agentic systems.
• Experience building reusable agent platforms, harnesses, or tool ecosystems.
• Experience with cybersecurity automation or security-sensitive systems.
• Experience operating open-weight models and commercial foundation-model platforms.
• Experience with enterprise-scale distributed or cloud-native systems.
• Ability to communicate system architecture, technical tradeoffs, and design decisions clearly.

Responsibilities

• Design and develop scalable architectures for production-grade agentic AI systems.
• Build agent orchestration, multi-agent workflows, state management, memory, tool integration, and human-in-the-loop capabilities.
• Design reusable execution harnesses and services that separate AI reasoning and planning from deterministic execution.
• Develop model-agnostic interfaces supporting multiple foundation models, providers, and inference platforms.
• Build and integrate MCP servers, APIs, tool registries, skills, enterprise services, and data sources.
• Develop retrieval and context services, including RAG, hybrid search, vector databases, and agent memory.
• Build backend services supporting asynchronous and distributed agent execution.
• Design evaluation and testing capabilities for agent behavior, tool usage, task completion, security, reliability, and performance.
• Implement observability across agent execution, model interactions, tools, state transitions, and system performance.
• Apply security engineering principles including least privilege, sandboxing, authorization, secrets management, guardrails, and auditability.
• Deploy and operate agentic services using containers, Kubernetes, CI/CD, and modern cloud-native engineering practices.
• Evaluate emerging models, frameworks, protocols, and technologies while maintaining architectural independence from individual vendors or platforms.

Benefits

• A comprehensive suite of benefits supporting physical, financial, and emotional wellbeing for team members and their loved ones.
• Personal and professional development programs intended to help employees pursue career goals.

What This Job Offers

• Job Type: Full-time
• Career Level: Principal