When it comes to its customers, the National Archives and Records Administration knows it has only a moment in time to meet their needs.
Whether it’s an elementary school student or an experienced researcher, NARA is striving to provide that same level of experience and care to each of their customers when they are searching for documents.
Sheena Burrell, who was NARA’s chief information officer until Nov. 21, said the end goal is to make sure those moments that researchers, students or any citizen come to NARA matter and they find what they need.
To that end, NARA is leaning into both technology and process improvements.
Burrell, who left NARA to become the chief innovation officer at the Federal Deposit Insurance Corporation (FDIC), said there are several ongoing programs that will deliver a modernized user experience.
“One of the big things that we do is look at our different customer bases. We have them in different groups, like a researcher, an archivist, a K-12 student, someone who’s just a novice and who wants to come in for the very first time with us. We put them into different groups,” Burrell said at the recent ACT-IAC ImaginationNation conference in October. “We do those journey mappings to understand how are they meeting with us, how are they interacting with our platforms and our applications. We are looking at each one of their experiences and doing that journey mapping so that we’re able to connect with them the right way.”
One of the big technology improvements on tap at NARA is the application of artificial intelligence capabilities to several of its platforms.
Burrell said one of the first pilots with generative AI will be with documents and information from the Warren Commission, which investigated the assassination of President John F. Kennedy.
“We have our National Archives catalog, but one of the things from a customer perspective that we’ve heard a lot is that sometimes people are not able to find what they’re looking for. We go back to that K-12 example, a middle schooler versus a person who is a trained archivist or a trained researcher. Those are two different types of experiences, and we want to be able to cater to both. So we are looking at generative AI and we’re looking at semantic search,” she said. “We’re looking at Vertex [from Google] so that we’re able to use optical character recognition (OCR) for the data, index this data and its millions and millions of records to be able to find the information very, very quickly. With generative AI, the piece that we’re using is a generative summary, but we want to make sure that we have disclaimers that this generative summary does not represent our views. It could be inaccurate, and we’re putting those disclaimers out there so people that understand that this isn’t the National Archives, that this is some AI.”
NARA testing GenAI tool
Burrell said she expects NARA to launch the GenAI pilot in the coming months, maybe early 2025. The goal is to get feedback and continue to improve the tools.
Optical character recognition combined with AI opens the door to a lot of opportunities for NARA.
For example, NARA is using semantic search to help researchers find records pertaining to their topic based on key contextual words.
“The generative AI piece is where it does look at those records, it’s generating a summary of what the record information is, and then it also provides a reference back to the actual document. We are supplying a link to our National Archives catalog because it does give you one record, but in the catalog, it could show you the whole series of records,” Burrell said. “We have all of our use cases on Archives.gov, and you can see a list of all of our AI use cases. We’re looking at Freedom of Information Act (FOIA) and how to utilize AI for access, review and redaction. We’re looking at it for declassification. How can we review a document and see if it meets these certain standards? If we can declassify it, we’re looking at it for ourselves in terms of it as a workforce tool. This is utilizing it to create documents and emails to be able to respond faster. There’s additional use cases for AI within the National Archives, specifically for the catalog. This pilot that we’re doing with the Warren Commission records is a precursor to maybe how we might be able to you utilize semantic search with our Archives catalog.”
New CIO’s to-do list
For the new CIO, Gulam Shakir, who has been NARA’s CTO for the last four years, Burrell said security, the continued implementation of zero trust capabilities and the continued digital transformation journey.
“In terms of legacy debt or technical debt, it’s really around our infrastructure and making sure that we continue to improve our infrastructure. That’s something that our administration has been really good with in terms of making sure that we are upgrading our infrastructure and having it on refresh cycles,” Burrell said. “But it’s those legacy applications, that we are moving to the cloud, working with our business partners, and also our partners within industry, so that we’re building the right thing for our customers.”
One way NARA is modernizing is through a project that received funding from the Technology Modernization Fund. In May 2022, the TMF Board awarded NARA $9.3 million to upgrade the system that stores and processes records requests about veterans.
So far, NARA has received about $5.5 million from the board.
“It’s our case reference management system in terms of all the cases that come in from our veterans. And with that one, we had a Siebel-based system. We have multiple different servers, multiple different items to be able to house this information, to save the data and it was Oracle-based, and now we’ve moved that to a software-as-a-service platform, and we’re really proud of that,” she said. “It hasn’t gone live yet, but we already see the potential benefits of it.”
Copyright
© 2024 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.
