The agency is expanding an enterprise data platform to all department elements and increasingly using artificial intelligence in daily operations. #
The Energy Department estimates that Quanta, its internal enterprise data platform, and Joulix, a generative artificial intelligence suite of tools, will generate $74 million in annual operational savings, an Energy official said Thursday.
In implementing internal AI tools more generally, “we weren’t so concerned with ‘oh this is a new technology.’ It was ‘how do we use this to meet the Energy mission,’” Bridget Carper Arnone, deputy chief information officer for architecture, engineering, technology and innovation, said at GovExec’s Government Efficiency Summit.
The department plans to expand Quanta access to all 88 departmental elements by the end of the fiscal year, Carper Arnone said.
Energy partnered with Databricks to create Quanta in 2025 after the Office of Electricity was given three weeks to sift through a billion documents for compliance with an executive order, she explained.
The amount of data the office needed to go through was too large for Joulix to process. Quanta helped the Office of Electricity condense the 200 gigabytes worth of documents to four gigabytes in 12 minutes, she said.
The data processing “allow[s] us to leverage AI after we fix part of the data problem,” Carper Arnone said.
Quanta is currently used in 44 offices across the agency, with more than 500 data sources, and has helped the Energy Dominance Finance Office project how heat waves would affect electric grids, Carper Arnone said. Other DOE offices are eyeing Quanta to improve funding notification processes and correct cybersecurity vulnerabilities that have been exploited, she added.
Energy’s broadening use of Quanta has come as the department increasingly adopts artificial intelligence tools.
Joulix, which includes a generative AI chatbot known as EnerGPT, has 21,000 users across department elements, Carper Arnone said.
While Energy’s national laboratories have access to chatbots other than EnerGPT, employees at those facilities are using the Joulix-based tool to compile emails, draft responses and write one-pagers to their leadership. Besides EnerGPT, Joulix has been used to help write job descriptions, speed up contract awarding and help offices comply with executive orders and policy directives, Carper Arnone said, adding that the agency receives feedback from employees on new potential AI applications through a dedicated Joulix email.
“We do go, from a governance standpoint, and look at the risk versus the time frame, and then the reward,” Carper Arnone said.