Pricing and claims for GF Forsikring
GF Forsikring · Frontend Lead · 2025 – present
Frontend Lead for GF Forsikring's digital insurance applications: designed the new Price Calculator from the ground up, with ~90% of features covered by end-to-end tests, and lead the architecture of Digital Claims and its internal preview tool.
- ~90%of Price Calculator features covered by end-to-end tests
- 80%+unit-test coverage across the applications
- UATphases demoed to Product Owners with very few failure reports
Context
GF Forsikring is a Danish insurer. Its digital platform is a Turborepo monorepo of Next.js applications: the public website, Digital Claims, an internal preview tool and a Storybook design system, sharing UI components, types, design tokens and typed OpenAPI clients.
My role
I took over as Frontend Lead mid-project, under a fixed deadline. I own the frontend architecture and technical decisions and coordinate frontend delivery in a cross-functional team of five frontend and four backend developers, together with GF's own developers. I keep them close to the architecture and decisions so the eventual handover is easy.
Backend and API architecture belong to the backend lead; I work with that team on API contracts and flag issues that affect the frontend.
Price Calculator
A new application in the monorepo whose frontend architecture I designed from the ground up. Customers choose insurance products, answer product-specific questions and go through calculation flows that differ per product, ending with a price.
I challenged parts of the original customer journey that I believed could be better and reworked them together with the UX team. ~90% of features covered by end-to-end tests.
Digital Claims and an internal preview tool
Customers file insurance claims online in a flow the server defines question by question: dynamic question components, server actions for answers, uploads and completion, and a sign-in step.
Next to it, an internal preview tool lets management see how a claim flow will look for customers while it is being configured.
From request to release
- Estimates at four levels: project, epic, feature and task, with tasks split into frontend, backend, QA and deployment.
- Product Owner requests turned into refined, deliverable work with a clear implementation approach.
- Each phase demoed to Product Owners for UAT; those sessions have produced very few failure reports.
Quality
Unit-test coverage generally above ~80%. Vitest workspaces, Testing Library and Playwright run in Azure DevOps pipelines alongside Lighthouse and accessibility audits, SonarQube and container image scans.
Stack
- Next.js 16
- React 19
- TypeScript
- TanStack Query
- Zustand
- React Hook Form
- Zod
- Tailwind CSS v4
- Storybook
- Vitest
- Playwright
- Turborepo
- OpenAPI
- Azure DevOps