AI Scientist: Build Production-Ready Agentic AI in Sydney

Shape the architecture, the evaluation methodology, and the technical culture of NAB’s AI function. Partner with talented AI engineers, product leaders and domain experts to solve problems across the AI portfolio. This is a delivery-focused role for practitioners who thrive beyond experimentation – delivering business outcomes using advanced AI capabilities (LLMs, agentic systems) in a safe, repeatable, and production-grade outcomes at enterprise scale.

Our people are customer obsessed. They prioritise the needs and satisfaction of the customer above all else. Our mindset fosters innovation and creates strong, lasting customer relationships as we strive to be the most customer centric company in Australia and New Zealand. In the role of AI Scientist, you will join the AI team that’s doubling down on customer obsession.

Key Responsibilities:

• Develop the core components of NAB’s agentic AI products: implementing agent reasoning loops, tool use interfaces, memory systems, context management, and orchestration logic.
• Translate architectural designs into robust, tested, production-grade implementations, working across the boundary of data science and software engineering.
• Partner with Product, Engineering, and Platform teams to translate AI capability into measurable business outcomes.
• Design and implement prompting strategies, retrieval pipelines, and context assembly that determine the quality and reliability of agent behaviour.
• Embed rigorous evaluation strategies for probabilistic systems, including offline evals, online experimentation, etc.
• Ensure AI behaviour is explainable, auditable, and aligned with NAB’s Responsible AI principles.
• Implement strong observability across AI systems, behavioural drift detection, anomaly monitoring, and incident classification.
• Conduct rapid experimentation to test hypotheses about agent design: what works, what fails, and why; building the empirical evidence base that informs architectural decisions.

What You Bring:

• Strong experience delivering applied AI/ML systems into production, with exposure to LLMs and agentic architectures preferred.
• Hands-on expertise in building automated evaluation pipelines, runtime controls, and CI/CD integration for AI systems.
• Experience operating in cloud-native environments and collaborating in cross-functional product, engineering and risk squads.
• Comfort working across ambiguity balancing speed, safety, and value in complex enterprise environments.
• Strong foundations in statistics, probability, and experimental design: you can design rigorous experiments, interpret results critically, and quantify uncertainty.
• Familiarity with prompt and context engineering, retrieval augmented generation, tool use patterns, and agent orchestration. Direct hands-on experience is strongly preferred.

Why This Role Matters:

• Influence how AI is trusted, governed, and scaled across NAB.
• Shape NAB’s next generation of AI platforms by setting technical standards that endure beyond individual use cases.
• Combine deep technical craft with enterprise-level impact—delivering AI that customers and regulators can trust.

A diverse and inclusive workplace works better for everyone. We know that our people make us who we are. That’s why we have built a culture of equity and respect - where everyone feels valued and appreciated for being their authentic selves. In partnership with our multiple Employee Resource Groups (ERGs), we continue to foster an inclusive environment, where all NAB colleagues’ unique backgrounds and identities are understood, respected and celebrated. We are committed to providing an environment where you can work your way.

For details on the recruitment process, and accessibility, or alternate methods to apply for this role, please visit NAB Careers. To discuss adjustment requirements, please contact the NAB Careers team (please reference job number) or visit our Careers page through the link above for other contact options.

If you think this role is the right fit for you, we would love to hear from you. Please note candidate screening and interviews may be conducted prior to the closing date of the job advert. Unsolicited CVs from agencies will not be accepted.