Automotive claims provider sees 27% increase in delivery speed with AI-driven modernization

The client provides claims processing technology for the insurance and automotive sectors. The company was preparing to launch a new AI-driven claims capability and expand into the USA, Australian, and Japanese markets.

The client aimed to modernize three Java applications supporting claims processing for more than 300 partners in over 30 countries. The goal was to consolidate the technology stack on .NET. The initiative focused on creating a clear view of each application’s role and embedded business logic, establishing a stronger foundation for future development, and supporting regulatory requirements.

At the same time, the client wanted to develop an AI-driven claims capability for expansion markets. Achieving both goals required a structured modernization approach that could preserve critical processes while preparing the technology stack for new claims workflows and AI-enabled functionality.

N-iX used Cursor to reverse-engineer the business logic inside all three applications. The GenAI analysis auto-generated technical documentation the client had never had, giving the team a reliable map of what each application did before touching the code.

With that documentation in place, N-iX built AI-generated integration tests to establish coverage across the applications' key business logic. This step gave the team a safety net for validating behavior through the migration, since none of the original code had test coverage.

N-iX then migrated all three applications from Java to .NET 8 through an agentic modernization pipeline, preserving functional parity across the stack. The combination of AI-assisted reverse engineering, automated test generation, and agentic migration let the team complete the work without manual rewrites of undocumented logic.

Three months against an undocumented, untested codebase is a tight window. A 27% acceleration on top of that timeline points to a project that, without AI assistance, would likely have run closer to four months. That means the AI-driven approach gave the client back roughly four to five weeks it could put toward the USA, Australia, and Japan launches instead.

"None of the three applications had a test or a page of documentation. Normally that means months of manual discovery before you can touch the code safely. Once the AI-generated tests hit 80% coverage before migration even started, this stopped being a guessing game," said an N-iX engineer on the account.

The client came out of this with a modernized claims stack and a lower risk profile than it started with:

• Faster modernization: All three applications moved from Java to .NET 8 in three months;
• AI-driven acceleration: The AI-assisted approach delivered a 27% increase in delivery speed;
• De-risked migration: AI-generated tests covered 80% of key business logic before migration, closing the gap left by undocumented legacy code;
• Reduced regulatory exposure: Auto-generated documentation and functional parity testing addressed the compliance and maintenance risk of unknown business logic.