About the Role
We are looking for an AI Engineer to develop and optimise AI agents and LLM-powered applications that will help advance Accenture's Security Operations Centre (SOC) capabilities. Working closely with security engineers and SOC analysts, you will design, build and operationalise AI-powered workflows that integrate with security tooling and help accelerate threat detection and response. You'll bring strong AI engineering expertise, with hands-on experience building production-ready LLM applications, agentic workflows and RAG solutions.
Responsibilities
•Configure, evaluate and tune large language models for security use cases, including prompt engineering, system prompt design, retrieval-augmented generation (RAG) and tool/function calling.
•Design, build and maintain AI agents and agentic workflows using platforms such as Azure AI Foundry, integrating with security toolsets such as SOAR, SIEM and EDR.
•Implement safeguards to protect against adversarial inputs, including prompt injection and jailbreaks, ensuring AI systems meet robust security standards.
•Architect AI solutions that are modular, maintainable and built to scale, following secure development principles.
•Implement solutions in line with our AI governance framework, covering areas such as data handling, audit logging, human-in-the-loop controls and regulatory compliance.
•Monitor and optimise LLM API usage and infrastructure costs, implementing strategies such as caching, model tiering and guardrails to ensure solutions are economically viable.
•Work closely with key stakeholders, security engineers and SOC analysts to understand operational challenges, prototype solutions and iterate based on feedback.
•Produce clear technical documentation to support the effective ongoing management of developed capabilities.
Requirements
•2+ years of experience within AI/ML engineering.
•Hands-on experience building and deploying AI agents or LLM-powered applications in production environments, including appropriate observability and evaluation mechanisms.
•Practical experience with enterprise AI platforms such as Azure AI Foundry, AWS Bedrock, Google Vertex AI or similar.
•Strong understanding of LLM concepts and vector databases, including context windows, temperature and sampling parameters, system prompts and tool/function calling.
•Experience implementing RAG pipelines end-to-end, including data ingestion, chunking strategies, embedding model selection and retrieval tuning.
•Experience implementing parsing pipelines for tasks such as document processing and ingestion, including handling unstructured data.
•Knowledge of AI security risks, including familiarity with the OWASP LLM Top 10, and an understanding of how these risks can be mitigated.
•Experience working within enterprise governance frameworks.
•Proficiency in one or more programming languages, such as Python, and familiarity with common AI/ML libraries.
•Ability to build custom tools that can interact with, or be exposed to, LLMs.
•Strong understanding of REST APIs, with the ability to integrate AI solutions with third-party platforms and internal tooling.
•A collaborative and engaging approach to problem-solving, with a willingness to work as part of a wider team.
•Self-motivated, results-focused and pragmatic, with the ability to manage competing priorities and deadlines.
Nice to Have
•Previous experience within cyber security or security operations.
•Knowledge of enterprise security tooling such as SOAR, SIEM and EDR.
•Experience applying AI to security operations, threat detection or incident response.
•Experience deploying infrastructure-as-code.
Conditions
•Competitive salary and package (depending on level of experience).
•Any offer of employment is subject to satisfactory BPSS and SC security clearance, which requires 5 years continuous UK address history (typically including no periods of 30 consecutive days or more spent outside of the UK) at the point of application.
