1736 - Verizon 利用 watsonx.ai 生成 AI 代码运营效率提升 50%.pdf

1、Orlando,FLOctober 69IBM TechXchange 2025Kapil Rajyaguru,IBM,Principal CSMSunny Singh,Verizon,FellowVerizon achieves 50%operational improvement using watsonx.ai to generate AI-powered code-1736Agenda01020304050607Business Use Case OverviewSolution OverviewDemoValue PropositionChallengesNext StepsQ&AI

2、BM TechXchange|2025 IBM CorporationBusiness Use Case OverviewIBM TechXchange|2025 IBM Corporation3OperationalImprovements identified by automation50%Method of Procedure(MOP):Method of Procedure(MOP):A comprehensive,step-by-step guide outlining the procedure for deploying network infrastructure,servi

3、ng as the foundation for the Verizon network team to develop Python/Ansible code that automates network deployment.Business ChallengeBusiness ChallengeInefficient Traditional MethodsInefficient Traditional Methods Manually writing code for a few hundred MOPs Relying on developers to write quality co

4、deSlow ProcessSlow Process Took weeks to deliver a single MOP to productionLack of Standardization and Lack of Standardization and EfficienciesEfficiencies Hindered the clients ability to streamline MOP creation and deploymentSolutionSolutionwatsonx.aiwatsonx.ai and WxO deploymentand WxO deployment

5、Deployed watsonx.ai&WxO on IBM Cloud Pak for Data to leverage LLMs and agents to generate Ansible/Python codeData IngestionData Ingestion Load the instruction document to generate summarization and codeUser validationUser validation User to review the generated code and make necessary changesResults

6、ResultsOperational ImprovementsOperational Improvements 50%operational improvements achieved through IBM Watsonx.ai solutionsReduced TimeReduced Time-toto-DeliveryDelivery Time required to deliver MOPs reduced by approximately 50%50%MOPs now available for production within daysdays,not weeksIncrease