Senior Data Scientist - Financial Crime

 

• Sydney, Melbourne, Brisbane or Perth Location with Hybrid Working. 
• Lead complex fraud and financial-crime solutions end to end. 
• Stay hands-on, strengthen investigation outcomes and develop the technical judgement of others. 

 

What’s the role? 

As a Senior Data Scientist in Westpac’s Enterprise Functions squad, you will lead complex and material fraud and financial-crime data science work. You will take ambiguous problems through to adoption and measured outcomes, remaining hands-on while providing technical direction, review and mentoring. 

Your work will span transaction monitoring, suspicious matter investigations, scams, anti-money laundering and counter-terrorism financing (AML/CTF), entity resolution, typology discovery and network analysis. You will combine statistical and machine-learning judgement with anomaly detection, natural language processing (NLP), Generative AI, graph analytics, neural networks and graph neural networks (GNNs). 

This is a senior individual-contributor role (AA Advanced Specialist), not a line-management position. You will independently lead complex or high-impact use cases, coordinate contributors and own methodological direction, evaluation integrity and delivery. You will partner with the Lead Data Scientist on domain-wide standards and the highest risk decisions. 

The impact you can make 

• Protect customers and the financial system through better detection, prioritisation and investigation of fraud and financial crime. 
• Translate model performance into better alert yield, investigation effort, losses prevented and control effectiveness, while considering missed risk and customer friction. 
• Deliver valid, explainable and supportable solutions, choosing advanced methods only when evidence and operational needs justify them. 
• Raise squad capability through practical standards, rigorous review, mentoring and reusable analytical assets. 

What you’ll be doing 

• Work with financial-crime leadership, investigators and domain stakeholders to elicit requirements, propose solution options and agree on measurable outcomes including guidance through delivery, adoption and outcome realisation. 
• Lead integrated use cases combining transaction behavioural, customer and account data, entity resolution, network structure, text, rules and investigator feedback. Translate typologies into testable hypotheses, features, graph patterns, labels and alerts. 
• Select fit-for-purpose rules, statistical methods, classical machine learning, deep learning, graph methods, GNNs, NLP or GenAI. Establish credible baselines and challengers rather than defaulting to the most complex technique. 
• Own evaluation for rare, adaptive and weakly labelled events. Address temporal leakage, selection bias, delayed or incomplete labels and class imbalance through back-testing, out-of-time validation, segment testing and sensitivity analysis. 
• Assess graph construction, edge semantics, temporal validity, community and path behavioural, leakage, explainability and scalability. For NLP and GenAI, test factuality, grounding, completeness, consistency, bias, prompt injection, data leakage and human-review outcomes. 
• Design how detection connects to investigation: review queues, explanations, thresholds, deferral conditions, feedback and approved decision rights. Keep risk indicators and model-generated content distinct from verified facts. 
• Partner with AI/ML, data, platform and operations engineers to deliver reproducible, observable, secure and supportable solutions. Ensure code, transformations, prompts, graphs, models, evaluation sets and thresholds are versioned, tested and traceable. 
• Set model-quality criteria, data and concept-drift measures, monitoring thresholds and review triggers. Lead technical incident diagnosis, measure post-release outcomes and initiate recalibration, correction or retirement when conditions change. 
• Own assurance for material work across Responsible AI, Model Risk, privacy, security, record-keeping and AML/CTF. Lead fairness, explainability, proportionality and vulnerable-customer assessments, and produce evidence for review, audit and second-line challenge. 
• Escalate unresolved methodological, regulatory, conduct and customer risks. Partner with authorised business and risk owners on final investigation, regulatory-reporting and customer decisions, and with engineering specialists on platform and service operations. 
• Review code, models, prompts, graphs and designs; mentor Data Scientists and earlier-career practitioners. Build reusable typology components, feature libraries, graph patterns and evaluation suites, and stop or redirect work when evidence does not support it. 

What do I need? 

• Advanced practical expertise in statistics, machine learning and rare-event evaluation, supported by strong Python and SQL engineering skills. Evidence of independently delivering complex analytical or machine-learning solutions into operational use. 
• Advanced depth in at least two specialist areas, with working fluency across the remainder: anomaly and behavioural detection; graph analytics, network science, embeddings or GNNs; entity resolution and record linkage; NLP and GenAI; and neural networks for sequence or representation learning. 
• For NLP and GenAI work, practical understanding of information extraction, summarisation, agentic workflows, grounding, structured outputs and rigorous evaluation. You are not expected to have equal depth across every method. 
• Strong command of calibration, temporal validation, uncertainty, threshold policy, explainability and sensitivity analysis. Ability to identify subtle data, methodology, implementation and control weaknesses in others’ work. 
• Practical understanding of model monitoring, drift, experiment tracking, CI/CD, observability and controlled lifecycle management, with disciplined testing, version control and reproducibility. 
• Experience leading fraud and financial-crime solutions in a large, complex organisation at a scale comparable to Westpac, such as major financial services, telecommunications or a similarly regulated enterprise. 
• Demonstrated understanding of transaction monitoring, suspicious matter investigations, fraud/scam detection, AML/CTF, typologies, entity risk or investigation operations. Understand criminal adaptation, weak labels, investigator bias, alert fatigue and operational capacity. 
• Understanding of customer, account, transaction, alert, case, device and relationship data, including how lineage and quality affect investigative conclusions. Experience translating technical evidence into operational policies, thresholds or prioritised actions. 
• Working knowledge of AUSTRAC and AML/CTF obligations, Model Risk, Responsible AI, privacy, security, auditability and human accountability. Capability to engage credibly with external regulators on Westpac’s behalf, within delegated authority, explaining methods, evidence, limitations, controls and remediation. 
• Strong technical judgement, intellectual honesty and calmness in ambiguity. Challenge assumptions constructively, make uncertainty visible and balance innovation with customer impact and proportionate controls. 
• Ability to influence investigators, engineers, product teams and assurance partners without formal authority. Communicate complex trade-offs clearly, mentor generously and remain accountable for outcomes after delivery. 

What success looks like 

• Your solutions are adopted in investigation and monitoring workflows and deliver measurable, sustained fraud, and financial-crime outcomes. 
• Methodological decisions, evaluation evidence and controls withstand review; explanations do not overstate what the data or model can establish. 
• Monitoring and production evidence drive timely improvement, escalation or retirement, with customer and operational consequences kept visible. 
• Practitioners rely on your technical judgement and become more capable through your reviews, mentoring and reusable assets. 

Ready to build what matters? 

Apply now and show us what you have built, how you approached the problem, and what changed because your solution made it into the hands of users. 

To get started, simply click on the  or  button. Please note that application closing dates are subject to change so don’t delay your application! 

We’re all about creating a supportive and inclusive community. We welcome everyone – no matter your age, gender, background, or abilities. We also provide additional support to welcome our veterans, Indigenous Australians, and neurodiverse community.

If you need any adjustments during the recruitment process, you can find out more information and additional contact details by visiting the "People with Disability and/or needing Accessibility Requirements" page on our website.