Most finance professionals encounter machine learning through vendor demos or conference talks, which leaves a gap between awareness and actual working knowledge. This course addresses that gap directly, using cloud environments that mirror what banks, insurers, and asset managers run in production today.
What the course covers
Participants work through the core pipeline: ingesting financial time-series data, cleaning it, building predictive models, and deploying them on cloud infrastructure. Tools include Google Cloud Vertex AI and AWS SageMaker, with datasets drawn from credit scoring, portfolio volatility, and transaction anomaly detection. No prior ML experience is required, though familiarity with Python and basic statistics helps.
The course does not promise that graduates will immediately build production systems. What it does provide is a grounded understanding of where ML adds measurable value in finance, where it introduces new risks, and how to communicate model outputs to non-technical stakeholders. Regulatory considerations around model explainability are covered in a dedicated module.
Who typically takes this course
The cohort usually includes quantitative analysts, risk managers, and fintech product managers who need to evaluate or oversee ML projects rather than build them from scratch. Some participants are developers moving into financial domains. Both groups find the case-study format useful because it grounds abstract concepts in recognisable financial scenarios.
Technical requirements
A laptop with a modern browser is sufficient. Cloud sandbox accounts are provisioned for each participant at no additional cost. No local GPU or specialised hardware is needed.
Programme structure
Week 1 Financial data in the cloud: storage formats, APIs, and data governance basics
Week 2 Supervised learning for credit and risk: logistic regression, gradient boosting, evaluation metrics
Week 3 Time-series forecasting with cloud ML pipelines: feature engineering for sequential financial data
Week 4 Anomaly detection in transaction data: unsupervised approaches and threshold calibration
Week 5 Model deployment and monitoring on cloud platforms: versioning, drift detection, alerting
Week 6 Regulatory and ethical considerations: explainability frameworks, audit trails, and model documentation