Major project
Insurance
Insurance pricing model in production
Build a premium model and deploy it so every quote can be traced to a model version.
The brief
Regulators and auditors ask insurers to explain every price. Build a premium model, then deploy it with versioning, automated tests and monitoring, so every quote records which model produced it and the team knows when the model drifts. Good looks like: a reproducible pipeline, a quote API that logs model versions, CI with release gates, and a drift report.
What you will build
- Build frequency and severity models for the premium
- Track experiments and register models in MLflow
- Serve quotes from an API that logs the model version
- Add automated tests and a release gate
- Monitor drift in the customers being quoted
- Document the model for an auditor
Free, open data to build it with
Check each dataset's licence before you publish your work. Your mentor gives your team its own version of the brief (data, scope or region), so every team's project is its own.