As on few occasions in the past, we are experiencing a historic moment of intense technological development. The so-calledFourth Industrial Revolution, driven primarily byartificial intelligence,automation, and other advances in computing, presents us with an increasingly urgent scenario: the stock of legacy systems and the ever-growing need to modernize them.

The challenge is enormous and therefore exciting: according to areport by McKinsey and IT-CISQ, approximately 70% of the software in use at Fortune 500 companies is more than two decades old.

As time goes by, it becomes increasingly difficult to subject legacy systems toapplication modernization processes. In addition, in an increasingly digitized world with systems offering more and more features, users' demands forquality are rising dramatically. No wonder:AIpresents us with realities that are already integrated into our routine, as well as possibilities that will make people's lives even easier. In terms of changing reality, the popularization ofartificial intelligenceis comparable to that of digital microcomputing, the Internet, and smartphones.

This whole picture shows that the ability to modernize legacy systems is a competitive advantage in this increasingly fast-paced market. However, in detail, what is the scenario for these applications within corporations? And what are the biggest challenges?

Refining the concept and looking at inventory

The analysis of a legacy system is not only based on its technical or functional obsolescence and its importance to the organization. Inapplication modernization projects, when we are in the stage of characterizing legacy systems, we analyze them from several dimensions:

1. Application architecture→ we verify the complexity of the architecture, the existence, detail,qualitylevel of the documentation, and frameworks used
2. Languages → we analyze the programming languages and versions used to begin identifying those that can be used as replacements
3. People → we assess the availability of analysts, both within and outside Inmetrics, who are proficient in the languages used in the system, to begin projecting the project timeline.
4. Hardware → we check the level of connection between software and hardware and which paths to take, especially if, as a result of the project, a major change inapplication architectureor even infrastructure modernization will be necessary, such as migration to the cloud.
5. Financial → we calculate: the current costs of maintaining the system; the costs involved in modernizing it; and what will be spent to maintain it in the future, after the modernization process.
6. Organizational → based on all of the above dimensions, we assess the importance of the system, especially from a financial, time, hardware, and human resources perspective.

The result of the analysis gives us an indicator of how "legacy" the system is. To give an example, if its architecture is not complex, the documentation is very detailed, our teams have technical knowledge of the languages, and the current maintenance cost of the application is low, this system may be less "legacy" than others in the organization.

To get an idea of how legacy a system can be, let's take as an example one of the most complex inventories to restructure, that of the US financial system. It is a stock of applications dominated by mainframe infrastructure and the presence of Cobol, an acronym formed by the expression Common Business Oriented Language. According to the Institute of Electrical and Electronics Engineers, a global institute that is currently the largest technical community in the segment in the world,there are approximately 200 billion lines of Cobol code in operation worldwide, supporting 70% of global commercial transactions.

The resilience of theseapplicationsis one of the main factors preventing their immediate replacement. While cloud computing service level agreements promise uptime ranging from 99.9% to 99.95%, mainframes achieve 99.999%. In practical terms, this translates to a risk of up to 43 hours of annual downtime on one hand, versus less than 5 minutes of downtime per year on the other.

Cobol and other legacy systems have become a huge problem for the US financial system. However, they are not alone in facing this type of difficulty—the reality in Brazil is not very different. According toa report by Febraban, the Brazilian Federation of Banks, by 2023, only 10% of core banking operations had been migrated to the cloud.

Although responsible for a large portion of today's data traffic, financial system applications are just one of several sets of legacy systems. The cost of this inventory is high.

Legacy systems: today's main challenges

When we look at the financial dimension of legacy systems inventory, the enormity of the challenge becomes even clearer. According to theMechanical Orchard report, the global cost of keeping Cobol mainframes running reached $1.14 trillion in 2022. Areport from the same year by IT-Cisq points out that the accumulated technical debt for these legacy systems in the United States alone amounts to $1.52 trillion.

However, the issue is not just about values. The cost is not merely operational; it is primarily one of security.Legacy systems have three times more security vulnerabilities than their modern counterparts.

In addition to costs and security risks, there is a knowledge gap: as time goes by, the number of workers with knowledge of the languages that make up legacy systems is decreasing. Looking again at the US financial system, aCobol developer is currently 55 years old on average. These people will retire at some point, and the workforce for this service will become even scarcer.

Modernizing legacy systemsis a common type of project here at Inmetrics. To carry them out, we combine advanced application architecture and engineering practices withartificial intelligence, and whenever necessary, we use our cloud computing infrastructure to deploy the new versions of the applications.

If your company has a complex legacy system inventory to deal with,please contact us! A member of our team of experts can answer your questions and, from there, we can start modernization projects to gradually reduce your inventory.Click here to talk to us, we will be happy to discuss your legacy system challenges!