AI-augmented application modernisation: the new frontier for efficient and safe code migration projects

As many insurers are aware, IT application modernisation projects can be very time-consuming, expensive and risky, while offering benefits that are not readily appreciated by business stakeholders.

However, IT application modernisation projects can unlock significant digital transformation opportunities, as well as addressing compliance or IT security issues.

“Legacy software that had been developed ad hoc 10–15 years ago and was modified over time now runs the risk of becoming obsolete and out-of-support, making it hard to find providers able to maintain and evolve these applications,” explains Giuliano Altamura, Global BU FSI General Manager at Fincons Group and UK Country Manager.

“It may also be difficult to find professionals that are able or want to work on outdated technologies such as Visual Basic, ASP, old Java frameworks, Cobol or PL/I.” 

“Older systems also often have limitations in fully exploiting the value of cloud adoptions, hindering insurers to benefit from the agility, operational efficiency, cost optimisation, simple integrations and scalability offered in that environment. Moreover, these kinds of applications expose enterprises to serious compliance and security risks,” adds Altamura.

Until now, insurance companies, therefore, either avoided to modernise and transform their legacy estate or embarked in expensive and complex transformation programs, aimed at adopting market packages or at totally re-designing and re-building custom applications. But such programs also came with the complexity and risks associated with core systems transformation and data migration.

Thanks to the advent and evolution of GenAI technologies, a new innovative approach capable of reducing the costs and risks associated with these types of projects is now possible. This new semi-automatic approach enables the migration of legacy applications from old architecture patterns, frameworks and programming languages to state-of-the-art ones, whilst safeguarding processes, application flows, business logic and data models, thus avoiding the risk and complexity of data migration.

Such an approach is particularly apt when the identified modernisation strategy intends to preserve the business functionality and data, but it can also be applied as the first phase of an evolutionary transformation approach in order to get rid of older technologies and create the foundation for step by step complete functionality redesign and rebuild.

In fact, Large Language Models (LLM) can be exploited to analyse software code based on obsolete technologies, identify portions of consistent code and generate new code written with modern programming languages leveraging accurately selected prompts. 

This approach significantly reduces costs and speeds up projects, also facilitating the work of developers in the fine tuning/integration/testing of the code generated leveraging Gen-AI. 

“The same code can be migrated many times during the project, allowing for incremental adjustments, following an iterative approach based on continuous code check and refinement,” illustrates Riccardo Palombella, Director of Business Area of the FSI Business Unit at Fincons Group.

However in order to effectively reduce projects cost, complexity and risk Large Language Models (LLM) and available Gen AI tools should be augmented with a comprehensive set of tools able to address the key project concerns, like source code categorisation, source code tokenisation, advanced prompt engineering, target artefacts composition and quality KPIs monitoring. 

“Fincons significantly invested in the last 12 months in order to create an application migration engine able to address the above concerns and effectively exploit the potential of Gen AI,”adds Palombella.

“The quantitative benefits in terms of timing and costs to be gained from adopting this approach are not pre-determined and need to be evaluated for each specific project as they depend on several factors,” continues Palombella. 

“Such as the combination between the original program language and the migration target technology, the structure of the original application and the volume of the code lines that needs to be migrated. However, in worst-case scenarios the reduction in cost and time can be as much as 20%, while in optimal scenarios savings of up to 70% compared to the costs of a traditional projects can be achieved.”  

On top of the benefits mentioned above, a GenAI assisted approach also allows a business to define quality and success KPIs and to measure them throughout the entire duration of the project as part of DevOps processes for additional transparency.

To face these initiatives successfully, it is important to follow a structured method. In order to maximise the approach effectiveness and the potential cost saving, based on its extensive experience, Fincons proposes as a first step a broad existing application analysis, where both architecture, source code, applied patterns and best practices are assessed. 

Then, in a second stage, Fincons will identify together with the client the target application architecture and the best migration strategy for each artefact/source file, based on the assessment outcome. Once such assessment stages are completed Fincons will proceed with configuring and customising the application migration engine mentioned above and will begin an iterative and incremental application migration process, which will automatically migrate the source application.

“In the evaluation phase of these projects, many insurance companies raise issues connected to the intellectual property of their application code, over which they wish to maintain protection and confidentiality,” concludes Riccardo Vescovi, Technology Innovation Hub Group Director at Fincons Group. 

“For this reason, Fincons has chosen to adopt an agnostic approach towards GenAI models and tools. In fact, we can operate on reserved versions of these tools already adopted by insurance companies or, as alternative, we can offer our protected environment to insurers, ensuring extreme flexibility to meet client needs.” 

This is a Fincons Group advertisement feature

To discover more about Fincons Group, visit their website.

Sponsored