Agentic AI Security Engineer

Middleware Systems; City of Westminster; On-site; GBP 90,000 - 130,000; Full time

Job summary

Middleware Systems seeks an experienced Agentic AI Security Engineer to design, implement, and secure AI-powered autonomous agents and multi-agent systems. You will protect AI workflows, implement guardrails, and align with enterprise security standards.

Responsibilities include threat modelling, vulnerability assessments, IAM/RBAC enforcement, and secure data handling across AWS/Azure/GCP environments. Collaboration with AI and cloud security teams is essential to success.

Qualifications

• 5+ years in cybersecurity, application security, AI security, or related fields.
• Hands-on experience securing LLM applications and AI agent workflows.
• Proven ability to identify vulnerabilities and implement guardrails for AI systems.
• Experience with threat modelling, penetration testing, and red-team exercises for AI environments.
• Knowledge of security standards and governance applicable to AI/ML systems.

Responsibilities

• Design and implement security controls for Agentic AI systems and multi-agent architectures.
• Identify and mitigate AI-specific security risks, including prompt injections and data leakage.
• Implement security guardrails for LLMs, AI agents, RAG pipelines, and AI-powered apps.
• Secure agent-to-tool interactions, API integrations, and MCP servers.
• Perform threat modelling, pentesting, and red-team exercises for AI systems.
• Enforce IAM, RBAC, and least-privilege access for AI agents.
• Protect enterprise data with masking, encryption, and strict access controls.
• Develop monitoring, logging, and incident response for AI activities.
• Collaborate with AI/ML engineers and cloud/security teams to deploy secure solutions.
• Establish AI security policies, governance, and compliance controls.

Skills

AI security

LLM security

Application security

Threat modelling

Security testing

IAM and RBAC

Cloud security

Python scripting

OWASP Top 10 for LLM Apps

MITRE ATLAS

Agent frameworks

Education

Bachelor's or Master's degree in Computer Science, Information Security, or related field

Tools

LangChain

LangGraph

CrewAI

AutoGen

Semantic Kernel

OAuth 2.0

Kubernetes

Docker

Job description

We are looking for an experienced Agentic AI Security Engineer to design, implement, and secure AI-powered autonomous agents and multi-agent systems. The ideal candidate will have strong expertise in AI security, LLM security, application security, and threat modelling, with hands-on experience protecting AI agents, tools, APIs, and enterprise data from security threats. The candidate will be responsible for identifying vulnerabilities in agentic AI workflows, implementing security guardrails, and ensuring compliance with enterprise security standards.,

• Design and implement security controls for Agentic AI systems, autonomous agents, and multi-agent architectures.
• Identify and mitigate AI-specific security risks, including prompt injection, jailbreaks, data leakage, and unauthorized tool execution.
• Implement security guardrails for LLMs, AI agents, RAG pipelines, and AI-powered applications.
• Secure agent-to-tool interactions, API integrations, function calling, and Model Context Protocol (MCP) Servers.
• Perform threat modelling, vulnerability assessments, penetration testing, and red-team exercises for AI systems.
• Implement identity and access management (IAM), role-based access control (RBAC), and least-privilege access for AI agents.
• Protect sensitive enterprise data through data masking, encryption, access controls, and secure data handling.
• Develop security monitoring, logging, and incident response mechanisms for AI agent activities.
• Collaborate with AI/ML engineers, application security teams, and cloud security teams to implement secure AI solutions.
• Establish AI security policies, governance frameworks, and compliance controls. Strong experience in cybersecurity, application security, or AI security engineering.
• Hands-on experience securing LLM-based applications and Agentic AI systems.
• Good understanding of LLMs, RAG, AI agents, multi-agent frameworks, and autonomous workflows.
• Experience with AI security threats such as prompt injection, indirect prompt injection, model manipulation, and sensitive data exposure.
• Knowledge of frameworks such as OWASP Top 10 for LLM Applications and MITRE ATLAS.
• Experience with agent frameworks such as LangChain, LangGraph, CrewAI, AutoGen, or Semantic Kernel.
• Knowledge of MCP security, tool-calling security, API security, OAuth 2.0, and identity management.
• Experience with Python Scripting and security testing tools.
• Understanding of cloud security across AWS, Azure, or GCP.
• Knowledge of security standards and governance frameworks such as NIST AI RMF, ISO 27001, and SOC 2.

Preferred Skills

• Experience with AI red teaming and adversarial testing.
• Familiarity with AI guardrail and security solutions such as NVIDIA NeMo Guardrails, Lakera, Protect AI, or equivalent platforms.
• Experience securing vector databases, embeddings, and RAG-based applications.
• Knowledge of Kubernetes, Docker, DevSecOps, CI/CD security, and secrets management.
• Relevant certifications in cybersecurity, cloud security, or AI security.

Experience & Education

• Experience: 5-10 years in cybersecurity, application security, AI security, or related fields.
• AI Security Experience: Hands-on experience securing LLM applications or Agentic AI solutions is preferred.
• Education: Bachelor's or Master's degree in Computer Science, Information Security, or a related field.