2784 Bridging Hybrid and MultiCloud ML Ops with CP4D Banco do Brasil Case Study v01.pdf

1、#internaOctober 21-24,2024Mandalay Bay,Las VegasJair NunesBanco do Brasil,AI ManagerZalkind DantasIBM,Tech LeaderBridging Hybrid and Multi-Cloud ML Ops with CP4D:Banco do Brasil Case StudyBanco do BrasilsHybrid PlatformSession ID:2784#interna08/24US$3,3 billionexchangeratefinancial profit23 million

2、pix keys75 million customers3 data centers1st half of 20246000 brand and offices in 14 countries16 billion daily transactions90%of them are digital#internaA2B2 PlatformPast,Present and Future#internaPlay the role of a robust and innovative ecosystem of tools that supports the development of analytic

3、al solutions at Banco do Brasil.Analytics PlatformDATA INTELLIGENCEEnable the development of Artificial Intelligence solutions in order to optimize processes,boost business,contribute to innovation and improve customer experience.IA PlatformDSMLAssistantsIA GenerativaIA ServiceFast DataDSMLExperimen

4、tationDataVizBig DataArtificial IntelligenceMachineLearningDeepLearningAnalyticsBI-BussinessIntelligenceDataScienceA2B2 Platform+#internaAI model monitoring goes into productionChoosing CP4D as a cloud development treadmillCompletion of the first MVP with CP4DaaSAI platform goes into productionFirst

5、 models produced on the AI Platform go into productionModel validation engine MVP completionThe validation engine on the AI Platform goes into productionWe reached 15 models in productionjuly2023december2019july2020december2020july2021december2021july2022december2022Integration of CP4D on premise wi

6、th the A2B2 Platform pipelinejuly2024december2023Implementing automated low-risk model validation#internaCP4D as a ServiceWatson StudioWatson Machine LearningWatson OpenScale#internaA2B2 PLATFORMBB CloudData ScientistData EngineerPrompt EngineerDialogue Curato