U.S. Citizenship and Immigration Services maintain more than 100 million active records for individuals who have had contact with the agency within the last 12 years, a number which its acting chief information security officer calls “staggering.”
The agency has had the notion to go digital for many years, but at a certain point USCIS moved away from building one giant system first and opted to integrate its existing systems in smaller pieces.
“Probably four or five years ago, maybe, when we started adopting cloud we began to move into more [of] an agile framework in terms of our development processes,” CISO Shane Barney said on Federal Monthly Insights — Application Services Month. “We embraced DevOps and eventually DevSecOps, and those qualities began to manifest themselves and the ability to do these sorts of things because now instead of taking on the entire elephant we’re taking on little chunks at a time.”
Paper records bring on several logistics including storage and warehousing the information. Barney said USCIS is hoping its eProcessing approach will help change that.
“eProcessing isn’t about digitization. That is really critical,” he said on Federal Drive with Tom Temin. “We’ve always wanted to digitize. We do have a process for doing digitization but it’s really complex. It’s very challenging to do.”
Rather, eProcessing is about taking current paper-based immigration processes — of which there are many — and making them digital processes. Barney said the goal is not to convert existing paper forms into digital records. But once eProcessing is in effect, those relevant forms will no longer be on paper.
“eProcessing in terms of its goals is quite simply just move away from paper. That’s really the objective,” he said. “And frankly it’s going to work. This is the first time I’ve seen where I can actually say yes we’re seeing the progress, we’re seeing our goals being met, the strategy is solid.”
But the first challenge is the data — getting different systems, which are currently stovepiped, to communicate with one another. Some data, such as names, birthplaces, citizenships, relationships are shared across several USCIS benefits forms. But at some point, the data branches off into “decision processes” for particular benefits, he said.
Luckily, the agency has had some practice with eProcessing. Its Electronic Immigration System (ELIS) already does some of this work for those applying for citizenship by using a “modified waterfall approach.”
“ELIS didn’t have certain aspects of the enterprise approach that we needed. And so we created a series of domains … common to all aspects of immigration,” Barney said.
These include intaking applicants’ information, maintaining their accounts online, managing risk and preventing benefits fraud. Those domains were built into a JSON format for easy translation.
The exchange of data on the backend is being done with microservices and application programming interfaces, Barney said. USCIS wants to become more data-centric in order adjudicate people’s immigration benefits and cut down on the number of mistakes made when processing people’s information.
“The real key here is not so much that we’re developing new applications [it’s] that we linking our current ones together,” he said. “What the goal of eProcessing is to do is link them altogether but do it in a very efficient, clean manner so that we can still continue to modernize in the background without affecting the front end.”