Tech modernization takes both human and artificial intelligence

Technology modernization: Not only do many federal IT managers think about it, they also are constantly on the lookout for strategies, specific technologies and services that will get them there.

Taking step back, a smart strategy should start with a clear definition of technology modernization in the first place.

Here’s one: “It focuses on the refactoring or replacement of legacy systems and technologies that are depreciated, replacing them with new technology stacks that better serve the customer and agency’s mission.” That comes from Brian Reynolds, a partner at Guidehouse. He added that tech modernization is “not just about introducing technically feasible solutions or just automating processes, it’s really about providing experience excellence.”

Modernization therefore requires understanding the ultimate users’ needs, including unmet needs, and the challenges constituents have dealing with systems an agency might deploy.

A principal challenge to modernization, according to Guidehouse partner Arijeet Roy, stems from the cost of maintaining and operating legacy systems.

“A lot of effort goes into keeping the lights on in existing older systems,” Roy said. “Immense amounts of money and effort go into just making sure that things run and do not break.”

This is where commercial cloud computing can help, Roy said.

“What the cloud has enabled is to help businesses focus on the functionality they want and making some of the other activities in it more, I would say, easy to achieve,” he said. The cloud “is there to take care of rudimentary stuff so that I can focus on building the applications, building the solutions that I need for my business.”

Added Reynolds, “Cloud is the single biggest enabler of what we’re seeing today with technology modernization.” Besides scalability and many services that IT departments can offload, Reynolds said clouds also provide a resilient infrastructure crucial to support modernized applications.

Given the finite amount of money and time any organization can bring to modernization and application development, “cloud is an amplifier and accelerator,” Reynolds said. “Organizations are able through cloud to do so much more with the resources they have, and to maintain those resources to refresh and regenerate those resources; and not be left with depreciated tech after just a couple of years.”

Modernizing takes a team

Benefitting from the utility of the cloud takes a team that understands cloud computing and what it can do.

“I think the workforce question is one of the biggest questions we have to think about when it comes to taking advantage of modern technologies like cloud,” Roy said.

Developing the workforce has two parts, he said.

“One is internal, where our agencies’ teams have to upskill themselves to come abreast with these latest technologies, Roy said. “Second is to create the ecosystem of training, vendors and other participants who bring the best strategy to people who can deliver the work.”

Added Reynolds, “There’s a level of smartness required to effectively deliver technology modernization, that I think is unprecedented.” Besides people with up-to-date development skills, the agency needs “folks who understand … the environments you’re deploying on, how to take advantage of elastic resources, how to auto scale, how to make sure that the solution is resilient and takes advantage of those services that make it resilient.”

To the human intelligence needed to undertake tech modernization, agencies should consider adding artificial intelligence. Like cloud, Roy and Reynolds said, AI can augment what people do by taking over certain crucial, but routine time-consuming functions.

“Developer productivity can be absolutely be improved through the use of generative AI,” Reynolds said. “That is especially the case when we’re talking about jumpstarting or the draft writing of code, or the refactoring of code.”

But AI also has more basic application to software, he said, such as generating documentation, inspecting your code base or analyzing usage patterns or constituent sentiment data to help pinpoint where you might need to update or rewrite code.

Still another AI application in the software cycle is to examine what Reynolds called infrastructure observability and application telemetry data. The purpose here is “so that we can use the AI to recommend improvements to our resource configurations, or our service deployments.”

Roy characterized such AI applications as the low hanging fruit for development teams. He noted, code itself can serve as the language in large language AI models.

But, Roy said “the whole idea of generating code using code, which is AI, would be something that still needs a little bit of maturity.” The reliability of code-generated AI depends on both the complexity of the business process you’re coding, and on “the digital skills and the maturity of the organizations to use those AI tools responsibly.”

Ultimately, Reynolds and Roy said, technology modernization must take on an industrial quality. That is, it must become an ongoing, repeatable activity using standard processes.

“By industrialization, we really are talking about a factory concept,” Reynolds said. He said the modernization “factory” should employ known best practices such as Six Sigma and continuous improvement “that are aimed at reducing work-in-process, reducing inventory, improving throughput. This is always done by removing process variability.”

He added that Guidehouse is “dedicated to standardizing the way we go about technology modernization, removing the need for heroes on projects, improving the predictability of our delivery efforts, and improving the reusability of the components that we deliver.”

Listen to the full show:

Copyright © 2024 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.