Companies long postponed modernizing business-critical legacy systems because of the cost, complexity and risk involved in replacing them. However, rising customer expectations and artificial intelligence’s impact on the economics of software development are changing this approach. Bupa’s renewal of its My Bupa mobile app shows that a legacy system migration can be treated not merely as a technology change but as a business transformation.
Asifa Sherazi, CIO of Bupa Health Insurance, says the risk posed by technologies approaching the end of their lifecycle can grow silently before suddenly emerging. Migrating the app from Xamarin to native Swift and Kotlin increased its rating from 3.7 to 4.7. The crash rate perceived by users fell by approximately 24 percentage points on Android and eight points on iOS.
According to Infosys executive Sanjeev Tripathi, artificial intelligence is changing the economics of modernization by reducing the effort, risk and time traditionally required. At Bupa, combining artificial intelligence-assisted reverse engineering with forward engineering enabled the transformation to be completed approximately 60% faster than in the pre-artificial-intelligence era.
Sherazi and Tripathi emphasize the importance of preserving institutional knowledge, increasing teams’ capacity to adapt and creating a working environment that enables problems to be reported early. In the future, modern platforms are expected to form the foundation for personalized, predictive and artificial intelligence-powered experiences. Tripathi says artificial intelligence will become embedded in platforms, from design through operations.
Why it matters
This example demonstrates the need to assess the renewal of legacy systems not merely as an infrastructure project under the sole responsibility of technical teams, but alongside customer experience, institutional knowledge and the way teams work. For this reason, the development concerns not only companies with similar systems but also executives who make modernization decisions by weighing costs against operational risks. Since the findings on the impact of artificial intelligence on time and resource use were presented in content produced in partnership with Infosys, it remains unclear how these findings will be tested against independent comparisons. In addition, the impact of short-term implementation gains on the long-term reliability of systems, team alignment and the preservation of institutional knowledge has yet to be answered.