Industry Workbench/Common Data Model

1 Industry Workbench/Common Data Model CDM driven analytics Vision at MicrosoftPayal Tiwana| Principal Group PM | Azure GlobalVolumeVarietyVelocityreport struggling to become mature users of data*report data silos and data management difficulties as roadblocks** Harvard business Review (2019), Understanding why analytics strategies fall short for some, but not for othersAnalytics & AI is the #1 investment for business leaders, however they struggle to maximize ROIC loud analytics enables business transformationEngagecustomersTransform productsEmpowerpeopleOptimize operationsData & IntelligenceORBusinesses are forced to maintain two critical, yet independent analytics systems Data lakeData warehouseEase of useFast explorationQuick to startProven securityAirtight privacyDependable performanceData lakeData warehouseData is stored within internal and external silosMarketplaceTransactionsSocialIoTAdv ertisingMobile/WebData intelligence is difficult to realize without a harmonized data estate especially if data scientists and business analyst use different tools and Vision.

2 Enable Data Harmonization using a Common Data Model Common Data ModelThe key lies inbuilding a unified data platform using a common data Model that describes and governs that data and is accessible by all personas MarketplaceTransactionsSocialIoTAdvertis ingMobile/WebOur Vision: Enabling App InteroperabilityApp 1 App 2 App 3 App 4 App 5 App 1 App 2 App 3 App 4 App 5 App from bespokeAppsto plug n play Appsthat can be more universally deployed quickly and cheaply requires a common data schema that the apps understand MarketplaceTransactionsSocialIoTAdvertis ingMobile/Web Copyright Microsoft Corporation. All rights reserved. Comprehensive Common Data Models Provide information blueprints for your business , create a plug and play ISV ecosystem and enable data consortiums across Industries5,000 data entities and30,000 data attributes(all with clear business names & definitions)Common Data Models22-30 BusinessAreasperindustrymodelRetailConsu mer GoodsFreight & LogisticsAgricultureOil & & EntertainmentFinancial Services & InsuranceUtilitiesEnergy & Commodity TradingSustainability & Carbon & Clinical TrialsPurviewData produced by Customers& their ecosystem partnersData from data providers/aggregatorsMicrosoft First Party Apps ( , D365)

3 GenomicsPower Query / Power PlatformSynapse Analytical ComputeAI / MLIOT / Digital TwinsAzure Data ShareCross Industry Data Pools & Consortiums Azure Synapse AnalyticsUnified experienceAzure Synapse StudioIntegrationManagementMonitoringSec urityAnalytics runtimesSQLA zure Data Lake StorageAzure Synapse LinkAvailable nowFuture supportSQL ServerAzureMachine LearningPower BIOn-prem dataCloud dataSaaS dataStreaming data Copyright Microsoft Corporation. All rights reserved. Azure Synapse AnalyticsQuery and analyze dataT-SQL using both provisioned and serverless modelsApache Spark in Synapse for quick creation of notebooks with your choice of languageBuild end-to-end workflows for your data movement and data processing scenariosExecute all data tasks with a simple UI and unified environmentSynapse SQLA pache Spark for SynapseSynapse PipelinesSynapse Studio Copyright Microsoft Corporation. All rights reserved. Azure Synapse analytics + Common Data Model (CDM)Query and analyze dataT-SQL using both provisioned and serverless modelsApache Spark in Synapse for quick creation of notebooks with your choice of languageBuild end-to-end workflows for your data movement and data processing scenariosExecute all data tasks with a simple UI and unified environmentSynapse SQLA pache Spark for SynapseSynapse PipelinesSynapse StudioComprehensive datamodels providing information blueprints for customers to describe their data estate for analyticsCommon Data Models+Automated Transformations+Low-code/no-code database design Copyright Microsoft Corporation.

4 All rights reserved. Democratize Data and unlock innovationAcceleration to the cloudRisk reductionCommon language for the organizationBetter data Integration Data Pipeline AutomationISV ecosystemIncrease the speed with which data estates can be moved to the cloud and modernizedUse proven data models that are 85-90% complete off-the shelf and do not require a large, lengthy and risky effortEnable clear communication among stakeholders and power business glossariesAvoid information silos by building a unified data lakeEnable automation of data pipelines in a low code, no code manner(future release)Power an ecosystem of ISV solutions that interoperate with Microsoft s data models Copyright Microsoft Corporation. All rights reserved. Customer case studyone of Europe's largest grocery retailersThe company wanted to bring together product and ingredient data sets together with customer data sets to determine nutritional choices customers were making in order to be able to recommend healthier choicesBusiness scenarioBenefits of using Common Data Models Process automation Providing data service: department after department Master Data Management DataMart in minutes Code generation versus Code typingBusiness outcomes 80% reduction in time to set up end-to-end data pipeline Autogenerate data warehouse in 2 weeks instead of 6 months Hours to setup data pipelines instead of weeks $$$ savings due to use of ready-made Industry data models and data automation toolsDemoReference Architecture Copyright Microsoft Corporation.

5 All rights reserved. Purview with CDMA utomated data discovery, lineage identification, and data classification across on-premises, multicloud, and SaaS sourcesUnified map of your data assets and their relationships for more effective governanceSemantic search enables data discovery using business or technical termsInsight into the location and movement of sensitive data across your hybrid data landscapePurview FeaturesCDM EnhancementsProvide semantics traits and meaning to optimize data discovery and classificationProvide verbose relationship descriptors through Model definitionsProvide rich business descriptions for entities and attributes for inclusion into business glossaryProvide Model lineage for derived analytical data models