ZotGPT Chat 生成式 AI 技术深入探讨.pdf

1、By Chris PriceUCI OIT Artificial Intelligence and ZotGPTLead Software Solutions ArchitectZotGPT ChatGenerative AI Technical Deep DiveCOLLABORATION FOR TRANSFORMATIONLets start with your QuestionsAs you enter the room,think of why you came,what you desire to know more about.Were going to do it liveCO

2、LLABORATION FOR TRANSFORMATIONWho I am1.5 years working in AI6 years in Research Administration IT5 years in Student Life and Leadership ITPublished Password Management ResearchUC Irvine Student in Japanese and Comp Sci7 years in software development before UCICOLLABORATION FOR TRANSFORMATIONHighlig

3、htsParticipants will learn about the challenges and solutions encountered in adapting open source applications,backporting application code,and migrating databases for a large audience.Participants will learn about the different technological infrastructures employed in deploying the ZotGPT Chat,inc

4、luding leveraging cloud platforms(Amazon EC2,DocumentDB,Azure API Management),and understanding the CI/CD pipeline in Github Enterprise with Terraform and Ansible.Participants will learn about how the UCI OIT AI Workgroup facilitated transformative change through teamwork,addressing safety and secur

5、ity concerns,and planning for future product development.COLLABORATION FOR TRANSFORMATIONHow?Collaboration across a team of diverse volunteers,passionate about bridging the gaps and reaching all of our community with this transformative technology.COLLABORATION FOR TRANSFORMATIONOutline:User Submitt

6、ed QuestionsArchitecture ModelRate LimitsRAG UseBiggest HurdleData ManagementMeasure SuccessOperational EfficiencyUser SegmentationSelf ServiceSecurity of IPAgents and AssistantsResearch UseFine TuneSearching through dataContent AccuracyWhat would I change if I started overCOLLABORATION FOR TRANSFOR