AI Validation Lead: GenAI & ML Models

Commonwealth Bank; Sydney; On-site; AUD 180,000 - 280,000; Full time

Job summary

Commonwealth Bank is seeking a Principal Data Scientist to lead independent review and validation of ML and AI models, including GenAI systems. You will design validation plans, mentor data scientists, and challenge governance with credible risk opinions.

You will apply advanced Python, ML tooling, and GenAI knowledge to validate orchestrators, sub-agents, RAG pipelines and guardrails, while staying ahead of evolving AI risks.

Qualifications

• Extensive experience in developing, evaluating or validating ML/AI models, GenAI or statistical models.
• Independent judgement with the ability to challenge stakeholders and defend positions.
• Hands-on experience with AI coding agents and related tooling.
• Strong Python and ML framework expertise, including model evaluation across complex pipelines.

Responsibilities

• Lead independent review and validation of ML/AI models across the Group.
• Design validation plans for complex AI systems and mentor junior data scientists.
• Set and improve AI validation practices, tooling and standards; provide credible challenge to governance.
• Communicate technical concepts clearly to senior stakeholders and governance forums.

Skills

GenAI knowledge

Agentic systems

Python

Code navigation

Test harnesses

Communication skills

Leadership (technical)

Evaluation pipelines

Independent judgement

Education

Degree in Statistics, Data Science, CS, Actuarial Science or related field

Tools

PyTorch

AWS SageMaker

GitHub/CI/CD

LangChain/LangGraph/OpenAI Agents SDK

Pydantic

Job description

Commonwealth Bank is seeking a Principal Data Scientist to lead independent review and validation of ML and AI models, including GenAI systems. You will design validation plans, mentor data scientists, and challenge governance with credible risk opinions.

You will apply advanced Python, ML tooling, and GenAI knowledge to validate orchestrators, sub-agents, RAG pipelines and guardrails, while staying ahead of evolving AI risks.