高效地为人工智能准备数据 [THR1135].pdf

1、 Efficiently Prepping Your Data for AIErik Hobbie,Oracle Industry Data ExchangeJason Wisdom,rStar TechnologiesErik HobbieProduct Manager,OracleIndustry Data ExchangeJason WisdomAI Architect,rSTAROracle System Integrator2Copyright 2025,Oracle and/or its affiliates|Confidential:Internal/Restricted/Hig

2、hly RestrictedThe following is intended to outline our general product direction.It is intended for information purposes only,and may not be incorporated into any contract.It is not a commitment to deliver any material,code,or functionality,and should not be relied upon in making purchasing decision

3、s.The development,release,timing,and pricing of any features or functionality described for Oracles products may change and remains at the sole discretion of Oracle Corporation.Safe harbor statement3Copyright 2025,Oracle and/or its affiliates4Copyright 2025,Oracle and/or its affiliatesWhat if you co

4、uld take advantage ofALL crucial data across your companyleveragingthe latest generative AI intelligenceto improve and contextualizeevery employee and customer touchpointCurrent Roadblocks To Achieve In Production5Copyright 2025,Oracle and/or its affiliatesPROBLEM#1 SILOS+PERFORMANCEMassive amount o

5、f dataIn distinct systems or storageOnly In Table/Column FormatNot optimal for Gen AISurface only the right dataOnly for the targeted individual or thing(e.g.a Utility Account)No or low hallucinationPROBLEM#2 RELEVANCE6Copyright 2025,Oracle and/or its affiliatesAppsGen AI Pipeline With Real-Time Ret

6、rieval(RAG)FORdozens of API calls analyzing 1000s of recordsin near real-timeRAG simply WONT PERFORM+RELEVANCE IS CHALLENGINGChallenges with RAG7Copyright 2025,Oracle and/or its affiliatesEnergy&WaterData ExchangeBob Jones is a personand has a utility bill with an amount due of$360.93with a total ac