利用 Blue Polaris 的数据释放 AI 生产力.pdf

1、Unlocking AI Productivity through DataTim Davis,Executive Director,Data,Executive Director,Data and AI Architecture and AI Architecture WorldWideWorldWideRyan Trollip,CEO of Blue Polaris,CEO of Blue PolarisTodays talk1Setting your AI strategy in an evolving marketThe importance of a focused AI strat

2、egy starts with a use case that delivers ROI2Lessons from Blue Polaris making gen AI workWhat were learning from real,successful,mission critical businesses applying AI at scale3Managing AI across the enterpriseAgents+assistantsSmaller,open modelsAccess to quality dataGovernance of any AIDeploy on a

3、ny cloud2Customer experienceDigital laborIT operationsAI use cases that unlock distinct valueUSD 3 billion savedApplication modernizationApplication modernizationCustomer serviceProcurementCustomer/citizen serviceHuman resourcesMarketingSupply chainIT operationsReduced supply chain costs by USD 150

4、million67%content creation efficiency gain for supported asset typesImproved initial build productivity for Ansible Playbooks by 45%33%increase in agent efficiencyReduced time from rsum submission to offer by 75%Gained 30%productivityOver USD 40 million in cost savings and avoidance1,500 hours of ma

5、nual labor saved yearly800 hours saved each monthHow do you manage the complexity of AI across the enterprise?AssistantsAgentsGovernanceModelsDataUse casesPut AI to work with watsonx.IBM watsonx is a portfolio of AI products that accelerates the impact of generative AI in core workflows to drive pro

6、ductivity.AgentsThe Evolution of Assistants&AgentsTraditional AssistantsSingle AgentMultiple Agents Rule basedPredefined action pathsTask basedPerformance constraintsLimited controlDomain based Specialized agentsControlMultiple architecture optionsoptionsThree areas of agentic innovation Prebuilt ag