Software Modernization: from industry needs to developers' perception

The large majority of existing pieces of software in operation are long-living systems (a.k.a., legacy systems), which represent strategic value to companies. However, over the years, user requirements changed, technologies evolved, and new business models emerged, leading to changes of such systems. As a result of extensive maintenance and obsolete technology, legacy systems usually have decayed and degraded architectures. Consequently, any maintenance/evolution activities such as fixing bugs, adding a new feature, or keeping up with new trends (e.g., digital transformation) become extremely complex, time-consuming, and costly (e.g., the US government spent over $90 billion on IT in 2019, from which about 80% to operate and maintain legacy systems). To remain competitive, efficient, sustainable, retain value, and embrace digital transformation, companies must have their legacy systems modernized. Nowadays, a common modernization strategy is to move systems to the cloud using modular and highly-decoupled architectures (e.g., microservices). However, several challenges are faced by practitioners when planning and performing modernization. This talk presents industry needs, challenges, automated support (i.e., using AI), and the developers' perception on using automatically generated solution. The content is based on work in collaboration with an industry partner and many research collaborators, resulting and several studies, covering both empirical results and solution proposals. Additionally, existing limitations/gaps in the field and research opportunities are identified.