1、 2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.D A T 1 0 5-SBuild Resilient GenAI&Agentic Apps with Intelligent MemoryDavid JoyVirag TripathiBuild Resilient GenAI&Agentic Apps with Intelligent MemoryTodays Agenda
2、Trends&ChallengesGenAI with CockroachDBIntro to CockroachDBDemoChallenges For GenAI AdoptionLegacy Systems&Infrastructure70%of organizations operate on legacy systems,and 50%of AI initiatives faildue to integration issues with outdated architectureData Readiness&Quality78%of firms are unprepared for
3、 RAG,agentic AI deployment due to poor data readiness and systemsScaling Beyond Pilots95%of enterprises have GenAI in pilot or early rollout stagesBuilding AI Applications with Scale Designing AI systems that can efficiently scale with increasing data volumes and user demandsHigh Costs&ROI Uncertain
4、tySubstantial investment in compute infrastructure,data pipelines etc,unclear ROISkill Gaps&Change ResistanceAbout 60%of enterprises lack proper training in GenAIWhere does CockroachDB fit?PC Computing EraMobile Computing EraAI Integrated Computing Era1980s/90s2010s/20s2020s/30sPurpose of Databases
5、has remained largely the same,though the way they function/operate has changedTraditional DBNoSQL/Distributed SQL DBChallenges for GenAI InnovationExisting DatabaseNo Vector Support Scale StrugglesMigration to vector databaseJanuary 28th,2021Recording Breaking Volumes Scalability955M1.6B1.4B1.2BUser
6、 GrowthDatabasesScaleGeographical ScaleApril 4th,2025April 7th,2025April 8th,2025April 9th,2025April 11th,2025GamestopSqueezeResilient&ReliableData LossOperational OverheadFinancial LossMulti-Region or Multi-Cloud or Hybrid Geo-PartitioningMulti-RegionCloud/HybridDeveloper Flexibility JSONTransactio