AI Model Risk Scientist - Validation & Oversight

Macquarie Group; New South Wales; Hybrid; AUD 140,000 - 190,000; Full time

Benefits offered by this job

Wellbeing leave

Parental leave

Volunteer leave and donation matching

Hybrid and flexible working

Learning and development support

Job summary

Macquarie Group is seeking a Data Scientist in Model Risk Management to independently review and validate AI use cases across our global business. You will assess fit-for-purpose AI systems, ensure risks are understood and controlled, and influence model approvals and risk decisions.

As part of a collaborative team, you will design validation approaches, evaluate data and methods, and clearly articulate findings to technical and non-technical stakeholders.

Qualifications

• Self-directed learner with curiosity to stay ahead of AI risks.
• Hands-on experience designing and evaluating AI/Generative AI solutions.
• Experience with retrieval-augmented generation and LLM evaluation frameworks.
• Proficiency in Python and modern AI tooling, including CI/CD practices.

Responsibilities

• Independently review, challenge and validate AI use cases across the firm.
• Design and execute validation approaches; assess data, methods and outcomes.
• Communicate findings to technical and non-technical stakeholders and influence risk decisions.
• Collaborate with model developers while providing constructive challenge for safe AI adoption.

Skills

Python

GitHub

CI/CD

SQL

R

TypeScript

AI frameworks

Tools

LangChain

LangGraph

OpenAI SDK

Google ADK

Langfuse

DeepEval

Job description

Macquarie Group is seeking a Data Scientist in Model Risk Management to independently review and validate AI use cases across our global business. You will assess fit-for-purpose AI systems, ensure risks are understood and controlled, and influence model approvals and risk decisions.

As part of a collaborative team, you will design validation approaches, evaluate data and methods, and clearly articulate findings to technical and non-technical stakeholders.