Get a reply from this employer
Your resume and cover letter, tailored to exactly what they’re hiring for.
Macquarie Group is seeking an AI Engineer to lead risk and controls for AI solutions, foundation models and data platforms within Commodities and Global Markets. You will apply the bank's operational risk framework across the AI lifecycle, addressing data ingestion, model selection, deployment, access and monitoring, while challenging risks such as autonomous actions and third-party dependency.
You will partner with the business to design and scale preventative and detective controls, create
You will lead technical risk and control assessments for AI solutions, foundation models, retrieval-augmented generation, agentic workflows, multi-agent systems and enterprise data platforms and apply Macquarie's operational risk framework to AI, data and technology-enabled activities across Commodities and Global Markets, supporting the identification, recording, assessment and management of material risks.
You will assess risks arising across the AI and data lifecycle, including data ingestion and use, model selection, deployment, access, human oversight, monitoring, change and retirement as well as provide practical challenge on risks such as autonomous actions and tool use, prompt injection, model or data leakage, excessive permissions, unauthorised access, unreliable outputs, model drift, concentration and third-party dependency.
You will partner with the business to design and maintain proportionate preventative and detective controls, including identity and access boundaries, runtime guardrails, observability, evaluation, escalation thresholds, kill switches and auditable decision trails. You will identify manual or judgement-based controls and drive their conversion into scalable automated or agentic controls, with clear fallback, escalation and human-approval thresholds
You will establish practical risk telemetry, control indicators and scenario testing to surface emerging concentration, access, conduct, resilience and third-party risks and lead technical incident triage, root-cause analysis and remediation for AI and data events, translating findings into reusable control improvements. You will be required to provide credible challenge to the business, technology and vendors while enabling proportionate, risk-based adoption.
You will lead technical risk and control assessments for AI solutions, foundation models, retrieval-augmented generation, agentic workflows, multi-agent systems and enterprise data platforms and apply Macquarie's operational risk framework to AI, data and technology-enabled activities across Commodities and Global Markets, supporting the identification, recording, assessment and management of material risks.
You will assess risks arising across the AI and data lifecycle, including data ingestion and use, model selection, deployment, access, human oversight, monitoring, change and retirement as well as provide practical challenge on risks such as autonomous actions and tool use, prompt injection, model or data leakage, excessive permissions, unauthorised access, unreliable outputs, model drift, concentration and third-party dependency.
You will partner with the business to design and maintain proportionate preventative and detective controls, including identity and access boundaries, runtime guardrails, observability, evaluation, escalation thresholds, kill switches and auditable decision trails. You will identify manual or judgement-based controls and drive their conversion into scalable automated or agentic controls, with clear fallback, escalation and human-approval thresholds
You will establish practical risk telemetry, control indicators and scenario testing to surface emerging concentration, access, conduct, resilience and third-party risks and lead technical incident triage, root-cause analysis and remediation for AI and data events, translating findings into reusable control improvements. You will be required to provide credible challenge to the business, technology and vendors while enabling proportionate, risk-based adoption.