Senior AI Validation Scientist

CommBank; Sydney; On-site; AUD 150,000 - 210,000; Full time

Job summary

CommBank is seeking a Senior Data Scientist to independently review, challenge and validate AI use cases across the Group. You will assess suitability, data quality, methodologies and outcomes, and explain findings to technical and non-technical stakeholders.

Model approvals and risk decisions will draw on your analysis and judgement. You will collaborate with model developers, product owners and business stakeholders, while remaining independent, helping to build the team’s validation

Qualifications

• A courageous mindset and a proven record of self-directed learning.
• Close attention to detail and a knack for spotting problems.
• Working knowledge of GenAI and agentic architecture, including orchestrators, sub-agents, and RAG pipelines.
• Hands-on experience designing, building and evaluating AI/Generative AI solutions, including synthetic data generation.
• Experience with GenAI frameworks and tools like LangChain, OpenAI Agents SDK, LangGraph, Pydantic, Langfuse/HoneyHive.

Responsibilities

• Review, challenge and validate AI use cases across the Group.
• Design and carry out validation for data, methodology, assumptions and outcomes.
• Explain findings clearly to technical and non-technical stakeholders.
• Support model approvals, usage decisions and risk acceptance with analysis and judgement.
• Collaborate with model developers, product owners and business stakeholders while staying independent.

Skills

GenAI concepts

AI model validation

Python

Git & CI/CD

Communication

Data analysis

Tools

LangChain

OpenAI Agents SDK

LangGraph

Pydantic

Langfuse/HoneyHive

GitHub

SQL

R

TypeScript

Job description

CommBank is seeking a Senior Data Scientist to independently review, challenge and validate AI use cases across the Group. You will assess suitability, data quality, methodologies and outcomes, and explain findings to technical and non-technical stakeholders.

Model approvals and risk decisions will draw on your analysis and judgement. You will collaborate with model developers, product owners and business stakeholders, while remaining independent, helping to build the team’s validation