AI Security Operations Lead for Agentic Platforms

AstraZeneca; Macclesfield; On-site; GBP 90,000 - 130,000; Full time

Job summary

AstraZeneca is seeking an hands-on technical leader to secure and scale an enterprise Agentic AI ecosystem, protecting data and accelerating science. You will own the AI security control plane, harden agentic systems and build production-grade automations across their lifecycle, guiding roadmaps as platforms evolve.

You will operate at the intersection of security operations, agent engineering and intelligent automation, delivering secure patterns, telemetry, and guardrails that enable

Qualifications

• Bachelor’s degree in Cybersecurity, Computer Science, Software Engineering, Data Science, Information Systems or a related field, or equivalent relevant experience
• Proven experience in Cybersecurity engineering, security operations, automation, platform engineering, AI/ML engineering or a closely related field, including meaningful hands-on work securing or operating Generative AI or Agentic AI solutions
• Proven ability to build production-grade automation or agents using Python; experience with APIs, event-driven workflows, testing, source control, CI/CD and operational support
• Practical knowledge of agent architectures, LLM applications, RAG, tool calling, memory, orchestration, multi-agent patterns, MCP/A2A concepts and associated security risks
• Hands-on experience with security monitoring, detection engineering, incident response and automation using SIEM/SOAR and relevant cloud, identity, endpoint, data or application security telemetry.

Responsibilities

• Agentic AI Platform Security and Control Engineering: Define, implement and continuously improve controls for agent identity, authentication and authorization; the least agency; tool access; data boundaries; memory governance; inter-agent communications; runtime policy enforcement; human approval gates; auditability and emergency containment.
• Secure Patterns and Guardrails: Create reusable secure patterns, reference configurations, guardrails and policy-as-code that teams can adopt without redesigning security for each use case.
• Architecture Assurance: Assess agent architectures and workflows; embed security into onboarding, design reviews, solution blueprints, threat models, release gates and production readiness; validate controls through testing, abuse-case analysis and red/purple-team exercises.
• Agent Lifecycle Operations: Own or coordinate secure lifecycle from discovery and registration through build, test, approval, deployment, monitoring, change, suspension and retirement; maintain inventory of owners, purpose, autonomy, privileges, data access, dependencies, risk tier and control posture.
• Workflow and Orchestration Reliability: Govern multi-agent dependencies, delegated actions, retries, approvals, exception handling and fail-safe behavior; drive secure, supportable integration of agents with enterprise platforms and Cybersecurity services via governed APIs and service identities.
• AI Security Observability: Build telemetry strategies that reconstruct agent intent and actions—including prompts and instructions where policy permits—tool calls, identities, memory updates, policy decisions, outputs, errors and outcomes; integrate with SIEM, SOAR, EDR/XDR, cloud, identity, data protection, application security and case-management platforms.
• Detection and Response: Design, tune and operationalize detections for AI-specific threats and control failures; lead or support triage, containment, eradication, recovery and post-incident improvement; codify automated response playbooks including safe pause, tool and credential revocation, network isolation, rollback, kill-switch activation, evidence preservation and escalation.
• AI for Cyber and Cyber for AI Automation: Lead the portfolio that uses agents to improve assessment, threat modeling, detection engineering, vulnerability analysis, incident response and reporting; embed security checks, risk scoring, control validation and approvals into AI/ML and agent delivery workflows; build agents and orchestrations in Python with secure engineering, CI/CD and production support; measure value through cycle-time, quality, coverage, analyst effort avoided, risk reduction and reliability
• Security Operations and Platform Stewardship: Provide technical oversight for day-to-day operation of AI security capabilities, ensuring health, telemetry completeness, integration fidelity, detection coverage and continuous service improvement; partner across SOC, cloud, identity, data, application security, DevSecOps, MLOps, platform engineering and architecture; deliver executive-ready dashboards and reporting.
• Governance, Risk and Regulated-Environment Assurance: Translate policy, regulatory and framework expectations into testable requirements and operational evidence; contribute to threat modeling and risk assessment using leading frameworks; support risk-tiered controls across research, business, regulated and GxP-relevant use cases; document residual risk, control limitations, exceptions and approvals to enable defensible decisions; engage vendors and peers to evaluate capabilities and shape adoption roadmaps.

Skills

Python

CI/CD

SIEM/SOAR

AI security

Cloud platforms

Education

Bachelor’s degree in Cybersecurity or related

Tools

CI/CD pipelines

Threat modeling

Job description

AstraZeneca is seeking an hands-on technical leader to secure and scale an enterprise Agentic AI ecosystem, protecting data and accelerating science. You will own the AI security control plane, harden agentic systems and build production-grade automations across their lifecycle, guiding roadmaps as platforms evolve.

You will operate at the intersection of security operations, agent engineering and intelligent automation, delivering secure patterns, telemetry, and guardrails that enable