By AI Trends Staff
The pandemic era of increased remote work and powerful available AI is motivating IT departments to examine legacy software systems for renewal. A legacy application, as defined by Gartner, is “an information system that may be based on outdated technologies, but is critical to day-to-day operations.”
This process of renewal can also be called modernization and often involves a move from on-premises hardware to the cloud.
“Getting rid of legacy is a perennial issue, but modernization is a top issue now more than ever,” stated Diane Carco, president and CEO of management consulting company Swingtide and a former CIO, in a recent account in CIO.
According to The State of IT Modernization 2020 report from IDG and tech company Insight, 26% of organizations are only at the beginning stages of IT modernization, and 19% have made moderate progress.
The first step of modernization is to take an inventory of what you have, Carco and other experts suggest. “Without that, you’re just going to spend a lot of money and have very little to show for it at the end of the day,” stated Thomas Klinect, a senior director and analyst with tech research and advisory firm Gartner.
Klinect suggests investing in an enterprise complexity analysis tool, to see how data is flowing through the organization. Then base the modernization strategy on the analysis, to reduce the risk of failure.
Citrix Making Incremental Approach to Modernizing
An incremental approach toward modernizing the application portfolio is well-advised, suggests Meerah Rajavel, CIO of Citrix, the software company with virtualization products.
“Think big, but deliver in increments,” Rajavel stated. “The idea of an ERP modernization can seem very daunting, because they are not pebbles to move, they’re mountains. If you try to move the mountain, you’re not going to know everything you need to know to move it when you start. It’s an impossible goal, and there’s a lot of risks. The execution has to be iterative, and when it is, the business gets value along the way.”
To modernize the company’s sales platform, which needed to continue to run during the process, she worked with sales and marketing leaders to determine their priorities. She delivered those first while determining next steps on the path to a fully modernized platform.
To ensure IT decisions have an impact on business value, experts recommend adopting practices of IT governance. The key stakeholders of the organization take ownership of the IT decisions, and IT management is responsible to these stakeholders. This can especially help when organizations move to a microservice architecture, in which applications are collections of loosely-coupled services. In what is also called a ‘polyglot’ architecture, decisions over which technology stack and programming languages to use are delegated to service providers, such as Amazon with AWS.
Rely on IT Governance Practices for Coherent Microservice Architecture
“But if you have dozens, or hundreds, or thousands, of microservices and each one is using its own blend of technologies, your IT infrastructure can get wildly out of control,” stated tech veteran Patrick Walsh, now senior vice president of training and technology with SkillStorm, an IT workforce development firm.
Walsh recommends the move to a polyglot architecture be made consistent with sound IT governance practice, which would include a process for selecting the technologies required for each service so that the environment does not become unwieldy and difficult to support. “It shouldn’t just be a free-for-all,” he stated.
Verizon considered the overall business strategy into the development of its IT roadmap and modernization plans, stated Shoma Chakravarty, vice president of enterprise architecture at Verizon. Any system that did not fit the business needs for agility, elasticity and reliability was not considered.
Verizon set priorities based on which systems would deliver tangible wins when modernized, leaving until later systems that may be older but are not creating a drag on business. For example, IT focused on upgrades to its UI/UX technologies because the work aligned with the company’s efforts to deliver a strong user experience, even though those technologies were not the oldest in the portfolio.
It’s best to think of modernization initiatives that do not have a start and end date, but are an exercise in continuous improvement. “Modernization is not… a point-in-time effort, it’s ongoing,” Chakravarty stated. “Technology is moving so fast, that even what we built yesterday is going to be due for modernization in the near future.”
Federal Agencies Identifying Systems to Modernize
The federal government is home to many legacy applications in need of modernization, according to the GAO. Some use out of data languages, some have unsupported hardware and software and many have known security vulnerabilities, according to a recent account in fcw.com. Many federal agencies are in the process of identifying systems that need modernization, and working through procurement decisions to ensure components meet regulatory and mission requirements.
Not every application, service, platform, or infrastructure is a good candidate for migration to the cloud. “It’s important to understand what the agency is currently running, where it is running, the associated service-level agreements (SLA), the uptime required for various workloads, and the sensitivity of those workloads,” stated Steve Thamasett, a senior security field solution architect at CDW·G, a company offering IT products and support for the education, government and healthcare markets.
AI and machine learning can be employed to analyze the data agencies need to run their systems, with the goal of saving money and improving services to citizens. AI can especially help to assess cybersecurity, how vulnerable the system is to attack. Advanced AI techniques aim to thwart hackers intent on breaching system defenses.
Agencies are finding many ways to use these technologies. The VA is incorporating AI in several projects to reduce wait time, monitor customer service and predict patient outcomes. The State Department is using data analytics and machine learning to better understand what is happening at posts throughout the world—and to better understand the technology that will be needed in the future.
IBM Offering Targeted System for Modernizing Legacy Systems with AI
IBM is offering a targeted service for modernizing legacy systems with AI, called Accelerator for Application Modernization with AI. It is a suite of tools within the Cloud Modernization Service designed to reduce the overall effort and costs associated with modernizing by tapping advanced AI technology from IBM Research. The accelerator leverages continuous learning and interpretable AI models to adapt to the preferred software engineering practices of the client. The new offering and service, that IBM is calling “AI for IT,” is the culmination of years of research and development at IBM Research into how AI can be used to transform the IT lifecycle.
The IT leaders investigating the move to a hybrid, multicloud architecture see that automation and AI are needed to help reduce IT costs and preserve business continuity, according to IBM’s AI blog. Many businesses have moved simpler workloads with lower complexity to the cloud, but they are struggling with the task of modernizing mission-critical applications because of incompatibility with cloud-native architectural principles. These applications span generations of software technologies—decades—and contain millions of lines of code with intellectual property and expertise locked in them.
IBM’s accelerator is meant to provide a guide. The new Watson AIOps services is said to “reimagine” IT operations with AI by examining patterns in IT operations, removing noise, correlating problems across multiple data sources and making recommendations to fix them. The Watson AIOps service can correlate among diverse data sources to localize the root cause of a problem, create an explainable diagnosis and recommend the best course of action, IBM states.
Read the source articles in CIO, fcw.com and IBM’s AI blog.