Your actuaries already know.
Your platform isn't telling the underwriter.
We help insurers translate actuarial knowledge into models that sit inside the underwriting workflow — explainable, governed, and trusted by the regulator.
Get an expert perspective
Average partnership: 5 years. We align our incentives with the loss ratio.
Four symptoms of an insurer with the right people and the wrong platform:
- 01
Underwriting — Loss ratios drift before anyone sees the pattern.
- 02
Pricing — Model recommendations land as friction, not guidance (no explainability in the workflow).
- 03
Claims — Document intake and routine adjudication consume the team.
- 04
Model Risk — Versioning and regulator documentation are reconstructed before each review, not produced continuously.
Four ways we work with insurers.
Each service is commissioned standalone or as part of a wider programme. Most clients start with predictive underwriting or claims automation and expand as the model-risk platform matures.
Predictive Underwriting
Risk models built with your actuarial team — calibrated, stress-tested, and surfaced inside the existing underwriting platform with an explicit rationale on every decision.
- ML architecture designed around the underwriting hypothesis
- Feature pipeline from your policy and claims data
- Models deployed into existing underwriting systems
- Explainable rationale at every decision point
- Regulator-grade documentation from day one
22% reduction in loss ratios; 30+ models in production.
Domain focus. Engineering in one team. Regulator-ready.
We start with the underwriting hypothesis and actuarial signal — not the technique.
Strategy and build in the same room. No handoff from advisory to implementation.
Explainability is a product requirement. Every score has a rationale; every decision traces back to policy.
Average relationship: 5 years. Incentives aligned with the loss ratio.
INTENT-DRIVEN SOFTWARE DELIVERY
Modernize business-critical insurance systems with AI-assisted, governance-first engineering.
AI proposes. Deterministic validators decide. Human governance remains explicit. Up to 70–80% faster delivery with traceability at every step.
The INTENTS Pipeline
Business requirements as structured, machine-readable specifications.
Shared vocabulary between actuaries, underwriters, and engineers.
Approach documented. Trade-offs owned and recorded.
Exact perimeter of change. No scope creep.
Machine-executable task list. Human-reviewable.
Non-negotiable engineering rules applied uniformly.
FINMA constraints, Solvency II rules — architecturally enforced.
Insurance Applications
Legacy Policy System Migration
COBOL, OpenEdge, LIFE/400 — extracted, mapped, governed, retired. 50+ years of historical truth preserved.
Underwriting Rule Formalisation
Actuarial logic extracted from spreadsheets. Encoded as versioned, auditable specifications.
New Feature Creation — Governed
Regulatory requirements delivered at 70–80% of traditional timescale. Full audit trail.
Behavioral Equivalence Protection
Deterministic validators prove behavioral equivalence. AI accelerates. Human governance is explicit.
From fragmented commission logic to a governed compensation capability.
Broker compensation sits at the intersection of distribution management, financial accuracy, compliance, and broker relationship quality. When the logic is opaque or distributed across systems, insurers face calculation inconsistencies, delayed payments, costly reconciliation, and reduced transparency under audit.
Qubiz implements a central compensation layer — decoupled from policy platforms and designed as a governed financial capability. This layer ingests policy and premium events from source systems and applies compensation rules through a deterministic engine with clear versioning, traceability, and historical protection.
The effective path is to externalise compensation progressively, while existing policy systems continue to operate. No large-scale cleanup required before starting. Compensation control from day one.
Capabilities
Consolidation across sources
Policy, premium, and business events aggregated across heterogeneous source systems into a reliable calculation basis.
Version-controlled schemes
Schemes managed with explicit validity periods, executed deterministically with transparent precedence and conflict handling.
Historical protection
Conditions valid at the time of policy signing remain preserved. New scheme versions apply to new business without corrupting historical correctness.
Broker mobility & continuity
Person-based, brokerage-based, or split entitlement structures — continuity maintained even when broker affiliations change.
Settlement, accrual & clawback
Controlled outputs for settlements, accruals, payment schedules, clawbacks, and reconciliation processes.
Full audit defensibility
Each result traced to scheme version, rule logic, triggering event, and calculation path. SLA control and exception governance built in.
What insurers gain from a governed compensation capability
Fewer calculation and payment errors
Deterministic rule engine — identical inputs always produce identical, explainable outputs.
Lower dispute volumes with brokers
Explainable calculations that can be shown to any broker on request, audit-defensible on demand.
Less manual reconciliation effort
Predictable outputs reduce costly reconciliation cycles and accelerate settlement workflows.
Stronger audit defensibility
Full lineage from policy event to compensation output. Regulatory-ready documentation produced continuously.
Management visibility into cost drivers
Analytics across brokers, products, channels, portfolios, accruals, clawbacks, and anomalies.
Faster, more predictable settlement cycles
Controlled, explainable, and strategically manageable — not a fragmented administrative process.
Commission calculations, scheme management, broker registry & entitlement control — operational today.
View Live PlatformFully decommission legacy platforms without sacrificing 50+ years of policy truth.
Legacy systems that hold inactive policies persist not because they are ideal, but because they are irreplaceable. Active policies can move. Inactive policies typically cannot — and that is why systems running on obsolete technologies, with shrinking SME knowledge, stay online. This is not a modernisation debate. It is a risk and continuity reality.
Closed policy lifecycle data is not operational — but remains analytically essential for long-tail claims analysis, actuarial modelling, regulatory reporting, audit traceability, and historical risk comparison. When this history is locked inside legacy platforms, analysts lose depth and decision quality degrades.
Using a modern lakehouse approach, inactive policies can be extracted, harmonised, and made accessible without rebuilding decades of legacy logic in new transactional systems.
Five-Phase Migration Approach
Legacy Discovery & Snapshot Extraction
Full snapshot extraction from aging policy administration systems, capturing all inactive policy records, contract terms, and lifecycle events without transformation loss.
Business Rule & Contract Recovery
AI-assisted analysis detects patterns in historical outputs, recovering undocumented contract logic and embedded business rules — including rules reachable only in the original contract text.
Schema Harmonisation & Canonical Mapping
Data from multiple legacy platforms harmonised into a unified canonical model with full lineage tracking from source to destination.
Immutable Layer & Time Travel Validation
ACID-compliant lakehouse architecture with time travel capabilities. Every record validated accurate before decommissioning begins.
Governed Retirement & Compliance Handoff
Controlled decommissioning with full governance, lineage, and regulatory compliance. Legacy systems retired. Knowledge preserved. Audit-ready.
Key Capabilities
Full Historical Preservation
Complete extraction without transformation loss. Every historical state, contract term, and lifecycle event captured.
Immutable Historical Layer
ACID-compliant lakehouse. No historical data accidentally modified or lost after decommissioning.
Time Travel & Versioning
Delta Lake version control and audit trails. Access any historical policy state at any point in time.
Unified Canonical Model
Data from multiple legacy platforms harmonised. Cross-system analysis without losing platform-specific detail.
Full Data Lineage
Every transformation logged, versioned, and auditable. Complete journey from legacy source to modern lakehouse.
AI-Ready Foundation
Modern lakehouse enables advanced analytics and AI-powered actuarial insights while legacy platforms are safely retired.
This is not modernisation for its own sake. It is controlled risk reduction through data preservation — transforming operational liabilities into governed analytical assets. Legacy systems can be retired while 50+ years of knowledge remains accessible, governed, and audit-ready.
Built for the Swiss regulatory environment.
FINMA, Solvency II, GDPR — every engagement in Swiss financial services and insurance carries compliance obligations that must be architecturally enforced, not manually checked. Our Swiss practice is led by someone who has served as Lead Architect across the Swiss insurance market.
FINMA Compliance by Design
Regulatory constraints encoded in the Safe Zone layer of the Intent Canvas — architecturally enforced from the first specification. Not added at delivery.
Swiss Insurance Domain Depth
Zurich Insurance, Helvetia, Swiss Re, Munich Re, Helsana. Marcel Ban — Lead Architect across Swiss financial institutions for 25+ years, based in Zug.
Multilingual Delivery
CH-DE, CH-FR, and EN across all delivery artefacts, stakeholder communication, and client-facing platforms.
Fixed-Price Entry Point
Swiss clients adopt on demonstrated results. Fixed-Price Discovery: your legacy system's business logic mapped in days. No commitment to what comes next.
Proven in Complex
Environments
Regulated industries. Long-term partnerships.
Results that appear in annual reports.
Predictive underwriting models that reduced loss ratios by 22%.
"Figura needed to move from retrospective risk assessment to predictive underwriting. We designed the ML architecture, built the feature pipeline from their policy and claims data, and deployed the models into their existing systems."
VIEW CASE STUDY
Our Clients
"Qubiz is a valuable IT partner, helping us solve all of our technical issues and challenges on all aspects."
Jeroen Klep, Product Manager Data & Analytics @ NEWCOLD
READ CUSTOMER STORY"Working with Qubiz felt like having a senior engineering team embedded in our own company from day one."
Remco Kuliman, Founder and Director @ AVICLAIM
READ CUSTOMER STORY"They understood our regulatory environment immediately. That defined everything for us."
Rhush Maugi, Co-founder @ FIGURA ANALYTICS
READ CUSTOMER STORY"Qubiz is a valuable IT partner, helping us solve all of our technical issues and challenges on all aspects."
Jeroen Klep, Product Manager Data & Analytics @ NEWCOLD
READ CUSTOMER STORY"Working with Qubiz felt like having a senior engineering team embedded in our own company from day one."
Remco Kuliman, Founder and Director @ AVICLAIM
READ CUSTOMER STORY"They understood our regulatory environment immediately. That defined everything for us."
Rhush Maugi, Co-founder @ FIGURA ANALYTICS
READ CUSTOMER STORY"Qubiz is a valuable IT partner, helping us solve all of our technical issues and challenges on all aspects."
Jeroen Klep, Product Manager Data & Analytics @ NEWCOLD
READ CUSTOMER STORY"Working with Qubiz felt like having a senior engineering team embedded in our own company from day one."
Remco Kuliman, Founder and Director @ AVICLAIM
READ CUSTOMER STORY"They understood our regulatory environment immediately. That defined everything for us."
Rhush Maugi, Co-founder @ FIGURA ANALYTICS
READ CUSTOMER STORYHead of Insurance Practice & AI Advisor
Speak with someone who has sat across from an underwriter.
Every insurance engagement starts with the underwriting workflow, the actuarial hypothesis, and the model-risk framework — not the technology stack. No discovery decks, just an honest conversation.
Start the conversation
25+ years across IT domains. DE · EN · FR
Credentials
- Lead Architect at Zurich Insurance, Helvetia, Swiss Re, Munich Re, Helsana
Ready to put actuarial expertise inside the underwriting workflow?
Fixed-price discovery. Four weeks. No commitment to what comes next.
Start the conversation