Oracle Analytics 在现实世界中的 AI 应用：一个相关的讨论 [BOF3470].pdf

《Oracle Analytics 在现实世界中的 AI 应用：一个相关的讨论 [BOF3470].pdf》由会员分享，可在线阅读，更多相关《Oracle Analytics 在现实世界中的 AI 应用：一个相关的讨论 [BOF3470].pdf（12页珍藏版）》请在三个皮匠报告上搜索。

1、Real-World AI Adoption with Oracle AnalyticsHow UF Health Jacksonville Transformed Its Supply Chain and Finance OperationsThomas Lee Oracle Analytics,ERP&SCM Sr.AnalystOracle AI World 2025|Venetian Expo 202|Oct 15,2025About UF Health JacksonvilleACADEMIC MEDICAL CENTER SERVING NORTH FLORIDA5,000+STA

2、FF|24X7 HOSPITAL OPERATIONS|MISSION-DRIVEN CARECORE SYSTEMS:ORACLE PEOPLESOFT+ORACLE ANALYTICS CLOUDDEPARTMENTS:PURCHASING,INVENTORY,RECEIVING,PAYABLES,FINANCEBefore OAC:The ChallengeFragmented data across ERP and SCMManual reconciliation,inconsistent reportingError ratio:11%Reactive problem-solving

3、;lack of predictive insightToo much data,not enough clarityOracle Analytics Cloud AdoptionImplemented OAC(20222023)for unified analyticsIntegrated PO,AP,Inventory,and GL data sourcesEnabled embedded AI and machine learningFrom reports to real-time insightsPurchasing:Data-Driven EfficiencyApplied the

4、 80/20 rule:top 20%of items caused 80%of errorsAI predicts mismatches before PO creationResults:Error ratio 11%3.9%,PO cycle time 27%Buyer efficiency increased across all unitsFinance:AI-Driven GL IntegrityAI pattern recognition on journal anomaliesDetected recurring posting&chartfield issuesAutomat

5、ed fix routines reduced manual review by 60%3 consecutive month-end closes under 2%errorInventory:Always-On Intelligence24x7 AI feedback monitoring demand and transactionsPredictive alerts for overstock/understockAuto reconciliation with PO and invoice flowsMachine learning retrains nightly on user

6、correctionsLeadership:End-to-End VisibilityUnified dashboard for executivesDaily visibility:trends,exceptions,initiativesTransparent culture:share both success and errorEvery error is a learning pointCulture:From 50:50 to 100%EngagementInitial adoption:50%coop