diff --git a/inventory/20260504-1.yml b/inventory/20260504-1.yml new file mode 100644 index 0000000..b3ab9c9 --- /dev/null +++ b/inventory/20260504-1.yml @@ -0,0 +1,59 @@ +template_date: 2026-05-05 +template_version: 1.0.2 +basic: + date: 2026-05-04 + version: 1.0.0 + contact_info: + - name: Sreenivas Karpurapu + division: DITD + mail: sreenivas.karpurapu@census.gov + - name: Ama Danso + division: DITD + mail: Ama.A.Danso@census.gov + +bedrock_information: + company_name: U.S. Census Bureau + website_url: https://census.gov/ + industry: Government + intended_users: + internal_employees: true + external_user: false + use_case_description: | + QUANTM's AI Enhanced Clerical Matching Solution evaluates the application of large language models to support + In Field Enumeration (IFE) and Census Data Quality Assurance (CDQA) by performing intelligent matching of complex records, + reducing manual clerical effort while producing transparent, explainable decisions consistent with expert human judgment. + +bedrock_models: + - anthropic.claude-4-5-sonnet + +federal_standards_information: + id: 20260504-1 + project_name: QUANTM + project_summary: | + As part of the quality operations of the US Census Bureau, computer matching and clerical operations are employed. These efforts help in the assessment + of the quality of data collection. The existing operations have constraints in their ability to accurately match and compare complex datasets, resulting in high number + of cases being flagged or removed. This process is time-intensive and expensive, creating a significant burden on human resources. With the advanced AI capabilities + available in the industry, QUANTM is assessing their feasibility to support the quality assessment efforts and reduce the clerical footprint. The bureau aims to explore + whether large language model (LLM) technology can replicate and improve upon the reasoning processes used by human clerks. + development_stage: active + data_sensitivity: | + Title 13. Demographic data from census survey responses + dms_project_number: 7535836 + cms_project_number: null + +deployment_details: + crf_number: NUMBER + finops_project_number: 1 + finops_project_name: edl_ditd + accounts: + - account_id: 001502248272 + account_alias: edl-addcp-prod-gov + regions: + - us-gov-west-1 + environment: prod + commercial_account_id: 563127383709 + commercial_account_alias: edl-addcp-prod-ew + +history: + - date: 20260505 + comment: created