The questions CIOs, CISOs, and QA leads ask before they sign.

Ratabase is where your rating logic is designed. Your core policy admin system is where that logic is executed on a live policy. There is always a risk of disconnect between the two — especially during new product launches or system updates. This gap is where premium leakage happens.

VeriQuant bridges the gap. It provides definitive, automated proof that your core policy admin system is perfectly executing the Workers’ Comp, Personal Lines, or Commercial Lines rules your business experts defined in Ratabase.

VeriQuant achieves this through its dual-mode architecture:

1. An Executable, Explainable Simulator. Your business analysts and underwriters author the complete rating logic — from Workers’ Comp to BOP — in a simple, version-controlled script that mirrors your Ratabase logic. The simulator becomes the executable source of truth, producing a step-by-step trace of every calculation.
2. A Live Execution Engine. VeriQuant tests your live policy admin system via its APIs and compares the final premium directly against the simulator’s result. Instant root-cause analysis of any discrepancy between Ratabase’s intent and the system’s execution.

For a carrier like yours, this delivers:

• Ratabase integrity. Auditable proof that the Workers’ Comp premium formula — payroll definitions, NCCI class codes, experience mod calculations — is 100% accurate from Ratabase to final premium.
• De-risked expansion. Validate BOP and Commercial Auto rating logic before launch, ensuring financial accuracy from day one.
• Faster speed-to-market. The simulator becomes the unambiguous “definition of done,” closing the gap between business and IT.

Yes. VeriQuant is built on Karate — the open-source testing framework embedded in Guidewire’s own testing toolchain for every Guidewire Cloud customer. Integration with PolicyCenter uses the standard Rating Service API and supports the full policy lifecycle (submission, quote, bind, endorsement, renewal).

Karate Labs is Guidewire’s first and sole testing technology partner. Upgrades to PolicyCenter pass through with no rework on the VeriQuant side — our integration is API-native rather than UI-scripted.

VeriQuant is engine-agnostic. It connects to any rating engine exposing REST, SOAP, or a message-based interface. For Duck Creek and Majesco, most deployments use their native APIs. For in-house engines, VeriQuant ships a thin adapter pattern that typically takes a week or two to configure.

We have reference integrations scoped across Guidewire, Duck Creek, and in-house platforms.

VeriQuant complements these — it does not compete with them. Your actuaries continue designing rating logic in their preferred pricing platform. VeriQuant validates that what was implemented in the core policy system faithfully executes that design.

The question is never “who builds the rating?” — it’s “did the rating the actuaries designed get implemented correctly?” Closing that gap is where VeriQuant lives.

No. VeriQuant exposes test runs via CLI and REST, so it drops into Jenkins, GitHub Actions, GitLab CI, Azure DevOps, or Harness with no custom plugin. Validation can be triggered on every commit, on a nightly schedule, or ad-hoc from the VeriQuant UI.

Run reports export to JUnit, TAP, CSV, and JSON — whatever your existing dashboards consume.

Local-first by design. VeriQuant runs entirely on-premise inside your VPC. There is no cloud offering, no multi-tenant SaaS, and no vendor-side telemetry. Rate tables, premium data, PII, and factor IP never leave your network.

This is a deliberate architectural choice, not a limitation. Enterprise carriers need to prove data residency and sovereignty to their CISO, their DOI, and their reinsurance partners — VeriQuant is built so those conversations are short.

VeriQuant validates rating in two modes simultaneously.

Execution mode calls your live rating engine through its API and captures the real-system result. Simulator mode runs the version-controlled script that mirrors the expected rating logic and produces a factor-by-factor trace.

When outputs match, you have high-confidence validation. When they diverge, the simulator shows exactly which factor, rating step, or surcharge caused the gap — typically pinpointing a defect in seconds rather than days of forensic work.

A plain-JavaScript scaffold that actuaries and business analysts author to represent the intended rating logic.

• Executable: it runs, it produces numbers, you can diff its output against the live engine.
• Explainable: it emits a factor-by-factor trace of every calculation step — no black box.
• Version-controlled: lives in git alongside the rating spec, reviewed by pull request, auditable by date and author.

The net effect: a 60-page Word spec nobody reads becomes a single, executable source of truth that developers build against and QA validates against. Your requirements are your tests.

A single server-side component running on commodity hardware. Typical mid-book deployment: 4 vCPU · 8 GB RAM · 50 GB disk. Containerized (Docker / Kubernetes) or bare-metal install.

No external database required — VeriQuant persists scenario sets and run artifacts to your existing Postgres, Oracle, or SQL Server. Stateless validation runs scale horizontally for very large combinatorial suites.

Not in any invasive way. VeriQuant talks to your policy admin system the same way your customer-facing portals do — through its existing APIs. No schema changes, no plugins, no triggers, no patches to apply.

It sits alongside your existing stack as a Platform Adapter. Your Insurance Platform stays exactly as-is.

No. VeriQuant has zero outbound calls by default. No phone-home telemetry, no usage analytics, no cloud dashboards, no license-server check-ins. Everything — test runs, scenario sets, validation artifacts, audit trails — stays inside your infrastructure.

Karate Labs has zero visibility into what your team does with the product once installed. That is a deliberate design choice to simplify CISO review.

Environment stood up in one day. First validation run on a PoV Line of Business within two weeks. Full production rollout across multiple LoBs typically runs 6–10 weeks depending on scenario-set breadth and state coverage.

The structured Proof of Value is fixed-scope: one LoB, one state, 8–10 weeks, $95K — credited against the implementation fee if you proceed.

Your operations team owns the server — same as any other internal service. We provide the platform, documentation, onboarding, and ongoing engineering support.

Karate Labs support covers incidents, upgrades, advanced scenario modeling, and integration patterns. We do not require remote access to your environment to support the product.

Yes — every validation run produces an audit artifact: scenarios executed, expected vs actual premiums, factor-level divergences, timestamps, and operator metadata. Artifacts export to CSV, JSON, and PDF for filing packets.

Retention is yours to configure. Artifacts live in your system for as long as your retention policy dictates — typically 7+ years for rating filing support.

All data stays inside your environment. VeriQuant doesn’t store or transmit PII beyond what the test scenarios you configure require. Data residency, sovereignty, and regional requirements (GDPR, CCPA, state-specific) are satisfied by the on-premise architecture.

No Data Processing Addendum is required beyond standard software licensing, because Karate Labs is not a processor of your data.

Because VeriQuant runs entirely inside your environment, it inherits your existing controls and certifications. We provide the platform; your CISO owns the perimeter. Your auditors treat VeriQuant the same way they treat any other internal business application.

Karate Labs’ own SOC 2 posture covers our vendor-side operations — build pipeline, release process, customer support systems — and is available under NDA.

Annual licensing by Direct Written Premium tier and scoped per Line of Business:

• Tier 1 (<$1B DWP) — first LoB $350K, each additional $250K
• Tier 2 ($1B–$10B DWP) — first LoB $600K, each additional $400K
• Tier 3 (>$10B DWP) — first LoB $1.5M, each additional $1.2M

Plus a one-time implementation fee ($250K) and annual maintenance (22% of license). Run the ROI calculator to model a 3-year TCO on your book.

Yes. A structured 8–10 week PoV, fixed at $95K, on one Line of Business and one state.

Deliverables: environment stood up in your network, scenario sets authored against your actual rate tables, a validated run against your live engine, and a formal validation report. The PoV fee credits against the implementation fee if you proceed to a full license.

Standard annual maintenance (22% of license) covers platform updates, security patches, and L1–L3 technical support. Premium tiers add scenario-modeling assistance, a named Customer Success Manager, and contractual SLAs for response and resolution.