Data Analytics for Banking Risk and Customer Insights Training Course

Course overview

Bank risk and customer teams hold large volumes of transaction, lending, behavioural and CRM data, yet often struggle to turn them into defensible decisions. Credit deterioration can be obscured by incomplete customer records and delayed indicators; operational losses may be buried in exception data; and customer campaigns can target the wrong segment when profitability, propensity and risk are analysed separately. This course equips participants to frame banking questions as analytical problems, build reliable datasets, and communicate findings that support risk appetite, portfolio action and customer strategy.

Participants work through the analytical workflow used in retail and commercial banking: profiling data quality, joining customer and account data, defining risk and customer measures, segmenting portfolios, detecting anomalous patterns, and building interpretable predictive models. They use SQL for extraction and transformation, Python for analysis and modelling, Power BI for controlled reporting, and SAS Viya concepts for governed analytical deployment. The course addresses credit risk indicators, expected loss inputs, arrears and vintage analysis, churn and next-best-action analysis, fraud and operational-risk signals, and model performance monitoring.

Delivery combines instructor-led demonstrations with realistic banking datasets and structured labs. Participants investigate a simulated bank portfolio containing loans, deposits, card transactions, customer interactions and delinquency outcomes. They produce a documented risk-and-customer-insight pack: a data dictionary, analytical segmentation, risk indicator dashboard, model evaluation summary, and prioritised actions for a risk committee or customer management forum. This gives managers a visible output that can be adapted to a live portfolio.

The course is suited to analysts who already work with banking data and need stronger methods for connecting risk, profitability and customer behaviour. It is particularly valuable where risk, finance, operations and customer teams need a common analytical language and clearer evidence for intervention decisions.

Course objectives

By the end of this course, participants will be able to:

• Profile banking datasets using completeness, validity, timeliness and reconciliation checks
• Write SQL queries that join customer, account, transaction and delinquency records into an analysis-ready dataset
• Calculate portfolio risk measures including delinquency rates, roll rates, vintage curves and exposure-weighted indicators
• Segment customers using behavioural, value, product-holding and risk attributes
• Build and evaluate an interpretable Python classification model for churn or early-warning risk
• Detect transaction anomalies using rule-based thresholds, peer-group comparisons and outlier methods
• Design a Power BI dashboard with risk appetite indicators, drill-through views and documented metric definitions
• Produce a governance-ready insight pack that states assumptions, limitations, actions and model-monitoring requirements

Benefits of attending

Target competencies

Who should attend

• Banking Data Analysts — who prepare portfolio data and need to produce risk and customer insight outputs
• Credit Risk Analysts — who monitor arrears, portfolio quality and early-warning indicators
• Customer Insights Analysts — who segment customers and assess churn, value and campaign opportunities
• Risk Reporting Managers — who need consistent metrics, traceable dashboards and decision-ready reporting
• Fraud and Financial Crime Analysts — who investigate unusual transaction patterns and prioritise alerts
• Retail or Commercial Banking Product Managers — who need evidence on customer behaviour, profitability and risk trade-offs

Requirements and prerequisites

Participants should be comfortable working with tabular data and basic descriptive statistics, including percentages, averages, distributions and simple charts. They should understand common banking terms such as customer, account, exposure, arrears, default, transaction and product holding. Prior experience writing simple SQL SELECT, WHERE and GROUP BY statements is expected; basic familiarity with spreadsheets or Power BI is helpful. No prior Python programming, machine-learning experience, SAS Viya licence, or advanced mathematics is required. Guided notebooks and starter code are provided, but participants should be prepared to inspect data and explain analytical choices.

Training methodology

Each day alternates focused instructor-led teaching with guided analysis of a simulated bank dataset. Demonstrations show the exact SQL queries, Python notebook steps, Power BI measures and SAS Viya governance concepts before participants apply them in individual and small-group labs. Case discussions require teams to distinguish a statistically interesting result from an action that is appropriate within risk appetite and conduct expectations. Daily outputs build toward a final risk-and-customer-insight pack, followed by a facilitated application-planning session for participants’ own portfolios, data sources and reporting cycles.

Course outline

Day 1: Banking data foundations and analytical framing

• Banking data domains: customer, account, product, transaction and risk-event records
• Analytical questions for credit risk, customer retention, fraud and portfolio profitability
• Entity resolution and customer-to-account relationship mapping
• Data quality profiling for completeness, validity, uniqueness and timeliness
• SQL SELECT, JOIN, CASE and GROUP BY patterns for banking extracts
• Data dictionaries, lineage notes and metric ownership
• Privacy, consent, purpose limitation and controlled use of customer data

Workshop: Participants create an analysis-ready customer-account extract in SQL and produce a data-quality scorecard with documented remediation priorities.

Day 2: Credit portfolio and operational risk analytics

• Exposure, utilisation, arrears, default and loss measures
• Delinquency buckets and migration matrix construction
• Roll-rate analysis for deteriorating lending portfolios
• Vintage curves and cohort comparison methods
• Concentration analysis by product, geography, sector and customer segment
• Early-warning indicators using payment, utilisation and behavioural signals
• Operational-risk event analysis and loss-frequency distributions

Workshop: Participants analyse a consumer-loan portfolio, build roll-rate and vintage views, and recommend accounts or segments for early intervention.

Day 3: Customer insight, value and behavioural segmentation

• Customer lifetime value and contribution-margin components
• Behavioural features from transactions, balances and channel interactions
• RFM scoring for recency, frequency and monetary value
• Rule-based and clustering-based customer segmentation
• Churn definitions, observation windows and label construction
• Propensity analysis for product uptake and retention actions
• Fair treatment and conduct-risk checks in customer analytics

Workshop: Participants develop a customer segmentation matrix combining value, engagement and risk, then define differentiated retention or service actions.

Day 4: Predictive models, anomalies and analytical controls

• Python pandas workflows for feature preparation and missing-value treatment
• Train-test splits, temporal validation and leakage prevention
• Logistic regression for churn and early-warning classification
• Confusion matrices, precision, recall, ROC-AUC and threshold selection
• Feature importance, explainability and adverse-action considerations
• Transaction anomaly detection using peer groups, z-scores and business rules
• Model monitoring for drift, stability, calibration and outcome tracking

Workshop: Participants build and assess a Python early-warning or churn model, select an operational threshold, and document performance and limitations.

Day 5: Decision dashboards and governed insight delivery

• Power BI star-schema design for banking portfolio reporting
• DAX measures for delinquency, roll rate, retention and customer value
• Risk appetite indicators, tolerance thresholds and exception views
• Drill-through design from executive metrics to account and segment detail
• SAS Viya concepts for model governance, approvals and monitoring
• Insight narrative structure for risk committees and product forums
• Action prioritisation, ownership and benefits tracking

Workshop: Participants complete and present a risk-and-customer-insight pack containing a Power BI dashboard, model summary, action register and governance notes.

Tools & standards covered

SQL, Python, Microsoft Power BI, SAS Viya

A typical training day

Live online deliveries follow the same structure in the East Africa Time zone, with shorter screen blocks and longer breaks.

What the fee includes

How you can take this course

Certification

Participants who complete the full five days receive the Skillset Development Certificate of Completion, stating the course title, course code, dates and delivery format — suitable for professional-development records and employer reimbursement.

Frequently asked questions