Establish PA Data Governance Council with Actuarial, Audit Ops, IT, Legal, and Compliance representation. Define data ownership model (PolicyCenter, NCCI, ADP as authoritative sources). Publish PA Data Dictionary v3.0 with 280-field taxonomy. NAIC, SOX, and state-specific WC regulatory compliance controls.
PA Data Dictionary v3.0Governance CouncilNAIC Compliance280-Field Taxonomy
Deploy NEXUS 4-zone Snowflake architecture: Bronze (raw ingestion, 4.8TB), Silver (cleansed, 3.1TB), Gold (audit-ready, 2.2TB), Platinum (AI/analytics, 1.8TB). Azure Purview lineage and Collibra data catalog. dbt 280+ transformation models with 340+ quality tests. RBAC and PII tokenisation.
Bronze → Platinum ZonesAzure PurviewCollibra Catalogdbt 280+ Models
MuleSoft API Gateway connecting all 24 PA data sources — 8 internal (PolicyCenter, BillingCenter, ClaimsCenter, Legacy Audit, Finance/SAP, HR/Workday, Salesforce CRM, Power BI) and 16 external (NCCI, ADP, IRS, OSHA, D&B, Experian, State WC Boards ×40, ISO/Verisk, ACORD, SOS ×50, BLS, FEMA, Paychex, License DB, Zywave, GL Rates).
MuleSoft API GW24 SourcesKafka Streaming4-Hour SLA
Great Expectations automated validation suite (340+ rules) across all 24 sources. MDM golden records for insured entity (D&B DUNS + IRS EIN), location, and broker. Target data quality score 61%→94%. Remediate top-20 critical DQ issues in EMod, payroll, and OSHA data impacting AI model accuracy.
Great ExpectationsMDM Golden Records61%→94% DQ Score340+ DQ Tests
Collibra enterprise data catalogue with 420-term business glossary specific to premium audit (WC class codes, EMod, OSHA TRIR, ISO GL rates, ACORD cert types). Azure Purview field-level lineage from source to Platinum analytics. Self-service discovery for auditors, actuaries, and data scientists.
CollibraAzure Purview420-Term GlossaryField-Level Lineage
24
PA data sources connected to NEXUS with validated dbt pipelines
72%
Data quality score by M6 — rising to 94% by M18
<30 min
Data latency SLA from source to Silver analytics-ready layer
NEXUS™ Technologies Deployed · Phase 1
❄️
Snowflake
Bronze/Silver/Gold/Platinum · 11.9TB · 152 tables · 99.99% SLA
🔗
MuleSoft + ADF
24 API connectors · Azure Data Factory ETL · Kafka events
🔄
dbt Core
280+ transformation models · 340+ quality tests · Nightly run
📖
Collibra + Purview
PA data catalogue · 420-term glossary · Field-level lineage
✅
Great Expectations
340+ automated DQ rules · Nightly validation · Scorecard reporting