Data Scientist (AI Engineering) - #1173014

Role Overview

The Data Scientist will support Responsible AI (RAI) within Data Strategy and Analytics Division (DSAD), combining technical evaluation of AI systems with the implementation and operation of agency AI governance processes. The role will assess model performance, methodology and risk, review policy compliance, and help project teams implement proportionate safeguards for safe and responsible AI adoption.

Key Responsibilities

• Evaluate models and methodology: Review AI use cases, data suitability, modelling assumptions, validation methods and performance measures. Conduct reproducible tests of accuracy, reliability, robustness, fairness and explainability, with findings documented for each assigned assessment.
• Assess AI risks and safeguards: Assess risk levels using applicable government and agency frameworks. For LLM and agentic systems, test relevant risks such as hallucination, prompt injection, sensitive data disclosure and unauthorised actions; recommend safeguards and assess residual risks.
• Process governance reviews: Review submissions for completeness and policy compliance, obtain supporting evidence, coordinate clarifications, prepare assessment recommendations and route cases to the appropriate approving authority. Maintain case records and track approval conditions and remediation actions.
• Implement policies and controls: Translate applicable IM8, Whole-of-Government RAI and agency requirements into practical checklists, assessment templates and workflows. Review and update these materials as requirements and use cases evolve, under DSAD’s direction.
• Develop evaluation and monitoring tools: Adapt reusable Python scripts, test sets and RAI tools with project teams and engineers. Define appropriate acceptance criteria and monitoring measures, and support reassessment following material changes or incidents.
• Report and transfer knowledge: Provide monthly updates on cases, risks, outstanding actions and delivery progress. Guide project teams on assessment requirements and hand over evaluation code, documentation and operating procedures to designated SWDA officers before contract completion.

Requirements and Qualifications

A. Educational Qualification and Experience

• Education: Degree in Data Science, Computer Science, Statistics, Mathematics, Engineering or a related discipline, or equivalent relevant professional experience.
• Experience: Minimum of 3 years of relevant experience in data science, machine learning or AI, including hands-on model evaluation and validation. Candidates should have demonstrated experience with end-to-end data science projects, assessing technical limitations and communicating findings to stakeholders.

B. Technical and Professional Competencies

Essential (Must Have)

• Strong technical understanding of LLMs, especially how they work, how to do prompt engineering effectively, and how to build apps with them.
• Good understanding of key AI safety and/or algorithmic fairness concepts, especially in how to operationalise and apply them to AI systems.
• Competent in Python programming, with some familiarity of cloud deployments for ML models and LLMs.
• Ability to apply RAI and risk assessment principles, assess policy compliance, document evidence and residual risks, and manage governance submissions and follow-up actions.

Preferred (Advantageous)

• Experience with LLM safety testing, prompt injection testing, fairness assessments, explainability methods or guardrails such as NeMo Guardrails and LLM Guard.
• Experience adapting RAI evaluation tools, building test datasets or integrating evaluations and monitoring into AI workflows with engineering teams.
• Familiarity with IM8 and the principles behind its requirements; ability to translate applicable requirements into practical checks and identify issues requiring escalation.
• Experience reviewing vendor-developed AI systems, cloud-based ML or LLM deployments, and supporting technical assurance or audit evidence.

C. Soft Skills and Communication Attributes

• Strong analytical judgement and attention to detail, with the ability to challenge assumptions constructively and distinguish evidence from unsupported claims.
• Clear written and verbal communication, including concise risk assessments and recommendations for technical teams, business owners and management.
• Ability to manage concurrent cases, follow through on outstanding actions and work independently while escalating material risks promptly.