2806 - IBM 数据智能代理：如何为企业用例构建 AI 代理.pdf

1、Orlando,FLOctober 69IBM TechXchange 2025Session 2806Jacob StellonIBM,Product Manager watsonx.data intelligenceLisa AminiIBM,Director Data&AI Platform ResearchIBM Data Intelligence AgentAI agents for enterprise use casesData Intelligence:an AI transformation from static memory to reasoning and action

2、Data Intelligence is the brain of the organizationWith AI agents its moving from memory,to reasoning and acting2IBM TechXchange|2025 IBM CorporationKey takeaways010203Data Intelligences AI transformationHow IBM Builds AI AgentsCo-create the agentic future with usIBM TechXchange|2025 IBM CorporationD

3、ata consumers struggle to find and understand data,so they rely on“data experts”instead of taking a self-service approachData practitioners find todays tools complex.Tasks require navigating heavy interfaces or depend on specialized skillsData teams need to federate governance responsibilities to da

4、ta owners,while keeping the process simple and consistent.IBM TechXchange|2025 IBM Corporation4What prevents converting data into actionable intelligence?Lack of self service consumptionTechnical overloadScaling data governanceConversational data discovery,lineage analysis,&data preparationIBM TechX

5、change|2025 IBM Corporation5Data Intelligence:an AI transformation from static memory to reasoning and actionAI Agents are more than chat6IBM TechXchange|2025 IBM CorporationAgents discover,prepare,package,and ensure data meets quality and governance standardsAgents write data quality checks and res

6、olve quality violationsWritten policy documents transform into executable data protection and quality rulesAgents dynamically identify gaps of data pipeline documentation and augment their resolutionIBM TechXchange|2025 IBM Corporation7Autonomous data product creationAutomatic data quality issue rem