Deep-dive guidewire policy data migration from PolicyCenter to Fusion + archive. Eleven domains pre-built: policies, terms, versions, risks, coverages, endorsements, premium transactions, party master, agency hierarchy, producer assignments, product catalog. Full version-chain preservation.
PolicyCenter's policy-term-version hierarchy is the most complex piece of any P&C migration. Every endorsement creates a new policy version with full premium recalculation, and the version chain has to survive intact for decades.
PolicyCenter models policies as policy-term-version hierarchies. A policy is the long-running master record. A policy term is the renewable period (typically annual). A policy version is created on every change within a term — an endorsement adding a vehicle, removing a location, changing a coverage limit, processing a mid-term audit, reinstating after lapse. Each version has its own full premium calculation, its own effective dates, its own rating-rule evidence, its own routing to GL as a written-premium adjustment journal. Statutory and regulator-facing audit requires the entire version chain for the policy's complete life plus retention horizon — often 7–30+ years depending on state and line of business.
A guidewire policy data migration that loses the version chain is worthless for statutory audit. The actuary cannot rebuild premium adequacy. The market-conduct examiner cannot verify rating-rule application. The reinsurance auditor cannot reconcile ceded premium to original policy basis. Syntra ETL preserves every policy version with full rating evidence, every endorsement transaction with before/after state, every audit-billing reconciliation with the final premium calculation. The downstream Fusion load surfaces version-level evidence per period; the long-tail archive preserves the full chain for the retention horizon.
Eleven core domains in scope, all out of the box: policies, policy terms, policy versions, risks (locations/vehicles/employees/dwellings per LOB), coverages (limits/deductibles per risk), endorsements, premium transactions (the audit-grade ledger per version per coverage), party/contact master (named/additional insureds, certificate holders, mortgagees), agency hierarchy (producers/sub-producers/agencies/MGAs), producer assignments per policy, product catalog (static product/rate per state per effective date) and rating-rule outputs (the calculated premium evidence).
Pre-built extractors and crosswalks for every PolicyCenter data domain. No bespoke CDA pipelines, no Gosu-aware unloads to engineer.
Every policy with full term and version chain. Effective dates, cancellation/reinstatement state, renewal lineage. Preserved end-to-end so any historical policy state can be reconstructed at any prior date.
Locations for commercial property. Vehicles for auto. Employees for WC. Dwellings for homeowners. Per-risk attributes, rating factors and underwriting evidence — multi-thousand-risk policies handled in parallel extraction.
Per-risk coverage limits, deductibles, optional endorsement schedules. Every endorsement transaction with full before/after policy state, premium impact and routing to Fusion GL as written-premium adjustment.
Audit-grade premium ledger per policy version per coverage. Rating-rule outputs captured as evidence metadata — what factor was applied, what tier was used, what override was approved, by whom, when.
Named insureds, additional insureds, certificate holders, mortgagees, loss payees. Producer hierarchy (agent/sub-producer/agency/MGA) with commission terms per LOB per effective date. Synced to Fusion Suppliers.
Static product definitions per state per effective date. Rating tables, classification schemes, form numbers, regulatory filing references. Preserved as substantiation for any future regulator recompute of historical premium.
A repeatable load order that respects PolicyCenter's version-chain integrity and Fusion's data dependencies. Typical timeline 8–12 weeks for the policy-data slice.
Inventory of every active product, custom Gosu rating rule, product catalog version per state per effective date. Volume estimate by LOB and policy version count. Audit endorsement scope mapped per LOB. Output: sized policy-data assessment with crosswalk design.
Product catalog extraction per state per effective date — the substantiation layer for every historical premium calculation. Crosswalks designed for premium-to-Fusion-revenue mapping per LOB per state. Reviewed by statutory accounting and underwriting leadership.
Named insureds extracted as Fusion Customers (FBDI Customer Import). Producers and agencies extracted as Fusion Suppliers (FBDI Supplier Import) with full hierarchy and commission terms. Loaded first as dependency for downstream premium and disbursement loads.
Full policy data extraction via CDA (GWCP) and/or JDBC (on-prem). Policies, terms, versions, risks, coverages, endorsements, premium transactions, rating-rule outputs. Staged as Parquet with hash-signed manifests partitioned by state and LOB.
PolicyCenter premium transactions converted to Fusion revenue journals via FBDI Journal Import. Written/earned/unearned split per LOB. Reconciled against PolicyCenter premium ledger to the cent. Version evidence carried per period.
Final delta replay via CDA incremental + Cloud API watermarks. Reconciliation sign-off pack. Long-tail archive sealed with state-level retention policies. Version chain queryable through archive UI for actuarial and regulator queries across the retention horizon.
Six structural advantages that compress the timeline and protect the audit chain.
Every policy version with full rating evidence carried through to Fusion (as version-level evidence per period) and to the archive (as the complete chain). No consultant scope-creep when actuarial discovers missing endorsement history three months in.
Cloud Data Access (CDA) extractors for GWCP and JDBC extractors for on-prem PolicyCenter run in the same pipeline. Hybrid insurers don't pay twice for two extraction profiles.
Commercial property with 500+ locations, WC with 30+ state employees, auto fleets with 5,000+ vehicles — parallelised per risk so extraction time scales with parallelism, not with policy size.
Every endorsement with before/after policy state, calculated premium impact, user, reason. Audit endorsements with final-premium reconciliation. Routed to Fusion as adjustment journals with full back-reference to source endorsement.
Full agent/sub-producer/agency/MGA hierarchy with commission terms per LOB per effective date. Synced to Fusion Suppliers in one move — eliminating the duplicate-supplier mess most insurers carry.
Static product/rate definitions per state per effective date preserved in the archive. Any future regulator or actuarial recompute of a historical premium has the full substantiation chain — no need to keep old PolicyCenter alive.
Guidewire policy data migration covers the full PolicyCenter data domain — policies, policy terms, risks, coverages, endorsements, premium calculations, written/earned/unearned premium ledger, party/contact master, agency hierarchy, producer assignments, product catalog and the rating-rule outputs that prove how each premium number was derived. It is far more than just 'move the policy table'. PolicyCenter's data model is built around the policy-term-version hierarchy where every endorsement creates a new policy version with full rating recompute, and the migration has to preserve the complete version chain for statutory and regulator-facing audit requirements that can stretch 7–30+ years depending on jurisdiction and line of business.
Eleven core domains, all out of the box. Policies (master policy records per LOB per state). Policy terms (the renewable period definition). Policy versions (every endorsement). Risks (locations, vehicles, employees, dwellings — LOB-specific risk units). Coverages (limits, deductibles, optional endorsements per risk). Premium transactions (the audit-grade premium ledger per policy version per coverage). Party/contact master (named insureds, additional insureds, certificate holders, mortgagees). Agency hierarchy (producers, sub-producers, agencies, brokers). Producer assignments (which producer wrote which policy). Product catalog (the static product/rate definition per state per effective date). Rating-rule outputs (the calculated premium evidence per coverage per version).
PolicyCenter creates a new policy version on every endorsement — adding a vehicle, removing a location, changing a coverage limit, mid-term renewal change. Each version has its own full premium recalculation, its own effective dates, its own rating-rule evidence. Statutory and regulator-facing audit requires the full version chain for the policy's complete life plus retention horizon (often decades). Syntra ETL's guidewire policy data migration preserves every version with its rating-rule outputs, effective dates, transaction-source-of-change (renewal, endorsement, audit, cancellation, reinstatement) and the user who initiated the change. The downstream Fusion-targeted load surfaces the version evidence per period; the long-tail archive preserves the full chain for the retention horizon.
Standard scope. A large commercial property policy can have 500+ locations with location-specific coverages, sub-limits and rating factors. A workers comp policy can have employees across 30 states with per-state rating and audit reconciliation. A commercial auto fleet can have 5,000+ vehicles with per-vehicle rating. Syntra ETL handles all of these — extraction is parallelised per risk, the policy-version chain is preserved per risk, the rating-rule outputs are captured per risk-coverage combination. Multi-TB volumes for very large commercial books are routine.
Endorsements are the operational heart of PolicyCenter and the most complex part of any policy data migration. Mid-term endorsements (add a vehicle, change a coverage, add a location) trigger a new policy version with full premium recalculation. The endorsement transaction records the change reason, the effective date, the user, the premium impact and the resulting policy-version state. Syntra ETL preserves every endorsement transaction with full before/after policy state, the calculated premium change, the routing to Fusion GL as a written-premium journal adjustment, and the audit-grade evidence trail. Audit endorsements (for workers comp and GL where final premium is calculated post-policy-period) are handled with the same rigor.
Critical for commission accounting and regulator-facing producer reporting. PolicyCenter's producer model is hierarchical — producers (individual agents) belong to sub-producers (offices), which belong to agencies, which belong to wholesale brokers or MGAs. Each level has commission terms, override commissions and split commissions on shared business. Syntra ETL extracts the full producer hierarchy, the commission terms per producer per LOB per effective date, and the producer assignments per policy. Producers and agencies are synced to Fusion Suppliers with tax IDs and 1099 classification preserved, eliminating duplicate supplier master and enabling Fusion AP to handle commission disbursements with the same hierarchy logic.
Different extraction paths, identical output. Guidewire Cloud Platform (GWCP) PolicyCenter is extracted via Cloud Data Access (CDA) for bulk Parquet of the policy data model, Cloud APIs for incremental delta access, and Cloud Studio query bundles for custom-rated product decoding. Legacy on-prem PolicyCenter (Tomcat/WebLogic + Oracle or SQL Server) is extracted via read-only JDBC against a read-replica with Gosu-aware metadata. Insurers running a hybrid (some LOBs on GWCP, others on-prem awaiting cutover) have Syntra running both profiles simultaneously without changing the downstream Fusion load.
Yes, with the standard parallel-run approach. Initial bulk extraction captures the policy state as of a fixed snapshot date. CDA's incremental Parquet exports and Cloud API modified-since watermarks then capture all subsequent transactions — new policy issuance, endorsements, cancellations, reinstatements, audits, renewals. Those deltas replay into Fusion GL throughout the parallel-run period. Live policy issuance, endorsement processing and renewal workflows in PolicyCenter are never disrupted. The cutover is at the downstream integration boundary — Fusion picks up its premium feed from Syntra instead of from the legacy nightly file drop — not at the PolicyCenter operational boundary.
Book a 30-minute discovery call. We'll walk through your PolicyCenter footprint — product catalog complexity, version-chain depth, multi-state mix, commercial-vs-personal lines split — and give you a concrete extraction plan, timeline and reconciliation strategy before the call ends.