Underwriting Analytics Reference Architecture for Insurance

This architecture demonstrates how insurers use the Databricks Data Intelligence Platform to analyze risk profiles for competitive rates and improved service quality. We showcase best-practice patterns for processing and analyzing operational and third-party data at scale with monitoring dashboards, alerts and quote recommendations.

Data flows

The following are descriptions of the data flows shown in the underwriting analytics architecture diagram:

1. Ingest internal/external (third-party) data either as batch or streaming from operational systems and ERP/CRM systems into the lake using Lakeflow Connect, Auto Loader and Declarative Pipelines. This helps to construct a holistic risk profile of individual or business entities seeking to be insured.
2. Lakeflow Jobs drives data transformation pipelines to join data, normalize across sources, derive analytical insights via data enrichment and populate Silver and Gold layers.
3. Interactively analyze insights in AI/BI dashboards or with natural language interaction in a Genie room. Rationalize any discrepancies via lineage tracing in Unity Catalog.
4. Apply actuarial models to get a baseline premium quote. The underwriter uses their experience to perform further risk analysis to get an optimum quote that is both competitive and aligned with the company’s risk appetite.
5. Serve up analytical artifacts and results via Databricks SQL to BI reporting systems, including General Ledger.
6. Visualize KPIs, tweak parameters to account for risk in Lakehouse Apps and share results securely with the operational system so the quote is passed to the concerned party.

Benefits

Benefits of using the Databricks Platform for the underwriting analytics architecture include the following:

• Establish best-practices architecture for underwriting analytics use cases
• Learn about automation available for underwriting teams, enabled by data intelligence