The US Army’s new generative-AI workspace enables users to apply large language models to daily tasks and boost productivity. But it takes a lot of compute power\u00a0and storage,\u00a0so the service has also been working to streamline how it handles cloud services and contracts.<\/p>\n
\u201cWe’ve been working to bring on something we call the Army Enterprise LLM Workspace, which is how we look at LLM usage safely at the [Impact Level 5]\u00a0for our warfighters\u201d and in daily office operations, Gabriele Chiulli, the chief technology officer for the Army\u2019s Enterprise Cloud Management Agency, told\u00a0Defense One<\/em>.<\/p>\n For example, LLMs were used to rip through the tedious task of reclassifying personnel descriptions, which includes defining and aligning job duties, experience, and backgrounds.<\/p>\n The Army was able to formulate about 300,000 personnel descriptions in a week using LLMs, Chiulli said. It would take a human about 10 minutes to review each description, which would amount to about 50,000 hours or 5.7 years of working around the clock.<\/p>\n Last year, the Army\u00a0previewed\u00a0plans for a generative AI pilot program. But the service had already\u00a0noted\u00a0a spike in cloud computing costs as more people used AI tools and services.<\/p>\n For example, an Army command was tasked with analyzing a problem set with national-level data within a 48-hour timeframe. They used generative AI to complete the task and it resulted in a big cloud computing bill and alerted the service to the need for \u201cguardrails\u201d to keep costs in check, Army tech leaders told reporters earlier this year.<\/p>\n \u201cHow we control costs is really important. Because what we certainly don’t want to end up with is a massive cloud compute bill,\u201d said Jennifer Swanson, deputy assistant secretary of the Army for data, engineering, and software.<\/p>\n In that cost-cutting spirit, the Army is simplifying its cloud contracting by consolidating services done by the same provider on the same contract. The service awarded Oracle \u201ca firm-fixed price task order through the U.S. Department of Defense (DoD)\u00a0Joint Warfighting Cloud Capability (JWCC) contract to provide the U.S. Army\u2019s Enterprise Cloud Management Agency (ECMA) with cloud compute and storage services,\u201d according to a news release.<\/p>\n Without everything on JWCC, \u201cit would be a square peg in a round hole to take an Oracle product and run it in a different cloud. Now, it works seamlessly, much more supported, much more secure when you’re running Oracle in Oracle, which we can do through the JWCC contract,\u201d Kim Lynch, Oracle\u2019s executive vice president for its defense and Intelligence business, told\u00a0Defense One<\/em>.<\/p>\n The move aims to streamline cloud management, enhance security, and reduce costs\u2014in some cases up to 50 percent, Chiulli said.<\/p>\n For example, one program\u2019s bill has been cut nearly in half.<\/p>\n \u201cWe’re talking [about] something that was $1 million a month, is now like $600,000 a month from a finance perspective because of some of the changes that we made,\u201d he said.<\/p>\n \u201cIf we have three separate contracts\u2026each one of these contracts has like one or two people that have to manage it\u2014all those different back-end processes,\u201d he said. \u201cNow, we only have to have one team’s worth of people to manage this stuff. So it really was about becoming really efficient from a personnel perspective to<\/p>\n","protected":false},"excerpt":{"rendered":" The US Army’s new generative-AI workspace enables users to apply large language models to daily tasks and boost productivity. But it takes a lot of compute power\u00a0and storage,\u00a0so the service […]<\/p>\n","protected":false},"author":1,"featured_media":2819,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[43,28],"tags":[87,77],"class_list":["post-2281","post","type-post","status-publish","format-standard","has-post-thumbnail","category-army","category-military-tech","tag-ai","tag-us-army"],"jetpack_featured_media_url":"https:\/\/thedefencenews.com\/wp-content\/uploads\/2026\/04\/US-ARMY-AI.png","_links":{"self":[{"href":"https:\/\/thedefencenews.com\/index.php?rest_route=\/wp\/v2\/posts\/2281","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/thedefencenews.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/thedefencenews.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/thedefencenews.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/thedefencenews.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=2281"}],"version-history":[{"count":3,"href":"https:\/\/thedefencenews.com\/index.php?rest_route=\/wp\/v2\/posts\/2281\/revisions"}],"predecessor-version":[{"id":2820,"href":"https:\/\/thedefencenews.com\/index.php?rest_route=\/wp\/v2\/posts\/2281\/revisions\/2820"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/thedefencenews.com\/index.php?rest_route=\/wp\/v2\/media\/2819"}],"wp:attachment":[{"href":"https:\/\/thedefencenews.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=2281"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/thedefencenews.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=2281"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/thedefencenews.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=2281"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}