Gradually renewing a 20-year-old policy system.
Agencies prepare quotes on one screen; the legacy core is modernised in stages, taking data and migration risks into account.
Problem and approach
Problem
At an established insurer, every policy passes through a single core system patched for twenty years.
In this scenario, 2,300 agencies prepare motor insurance quotes across nine screens, and the system closes for a weekend for every release. Most business rules exist only in the code and a few experts' memories; neither leaving the system untouched nor replacing it all at once is acceptable.
Approach
We plan to modernise the core through small, verifiable stages.
In a two-week discovery sprint, we extract business rules from code with domain experts and test a prototype with actual agencies in week three. The jointly scoped first phase, such as the motor insurance quote screen and event bridge, targets a launch in 6–12 weeks; migration of other products follows the roadmap.
One screen for the agency, supported by three components.
Quotes on one screen
A nine-screen flow becomes one form. Vehicle, coverage and package share one page; the premium updates with every choice.
Event-driven bridge
The old and new systems stay synchronised through Kafka events. Both sides see the same policy throughout migration.
Controlled release transition
Blue-green deployment simplifies traffic switching in this scenario. Rollback of data changes is planned and rehearsed separately.
Business rules catalogue
Rules extracted from code are documented in a readable catalogue and protected by tests. The impact of changes is visible before production.
Scenario and scope
Example users
Agency user, product and operations specialist, IT team and data owner.
Example workflow
The agency prepares a quote in the new interface; results are compared with legacy rules and the authorised team reviews differences. An approved module opens to pilot users; rollout expands as data reconciliation is completed.
Decisions for a real project
Define source code and data access, the first module, side-effect-free parallel testing, a maintenance window and how new records will be preserved during rollback. Zero downtime is an illustrative target for this concept, not a guarantee for every real project.
Related service: Software modernization. You can list your own users and systems in the project brief tool.
Timelines, teams and technologies in this scenario are illustrative planning choices; actual scope, schedule and team are defined during discovery. Product names do not imply partnerships.
What we measure
These targets are example measures defined for this scenario; actual results vary by project.
Technology and team
Technology
- .NET 8
- C#
- Azure
- Kafka
- React
- TypeScript
- SQL Server → PostgreSQL
For scope, data synchronisation and delivery terms when modernising a core system in stages, see the software modernization page.
Suggested team 7 people
- 1 solution architect
- 1 product designer
- 4 software engineer
- 1 DevOps engineer
Your process
Are you working with a similar system? Let's schedule a 30-minute conversation about modernising a legacy policy system while managing migration risks.
Let's discuss this conceptThis is an example, not a limit; we apply the same approach to similar processes in other industries.