Can you build the unified semantic layer my analytics, AI, and regulatory reporting all sit on top of?
Every carrier and large agency has the same problem: book-of-business truth is scattered across Applied Epic / Vertafore / Guidewire / mainframes / Excel, with inconsistent entity resolution between policy, client, location, and exposure. A modern lakehouse (BigQuery / Snowflake / Databricks) with deterministic entity resolution and a semantic layer is the precondition for everything else — analytics, AI, regulatory reporting, and M&A integration.
combined_policy_master_v2 at ~325K rows, book_of_business_v2, attrition_*, client_portfolio_v1, 10× fin_* daily-refreshing views, daily scheduled queries with validation runs. Full case study: Enterprise Data & AI Transformation at SWIA.| Quarter | Deliverable |
|---|---|
| Q2 2026 | Insurance dataset repository v0 — 30+ open datasets indexed |
| Q3 2026 | Insurance dataset repository v1 on Hugging Face + GitHub |
| Q3 2026 | Semantic-layer template (dbt + Power BI / Looker) |
| Q4 2026 | Public entity-resolution reference notebook |
| Q1 2027 | Speaking slot at one data-platform conference (Snowflake Summit / Databricks Data + AI Summit) |
Guidewire, Duck Creek, Applied Epic, Vertafore, Majesco, EIS, Socotra roadmaps · Snowflake insurance vertical, Databricks Lakehouse for Insurance, BigQuery insurance solutions · IVANS, AMS360, Sagitta, ACORD Forms 25/27/28 · dbt Labs, Atlan, Collibra, Alation, Monte Carlo case studies · NAIC, FEMA, SERFF, state DOI rate filings.
| For more details see LinkedIn | GitHub |