1、Orlando,FLOctober 69IBM TechXchange 2025Session code 2941Session code 2941Hailey Thao Quach,Joseph Santarcangelo,Faranak HeidariHailey Thao Quach,Joseph Santarcangelo,Faranak HeidariInformation Developer,Data Scientist,AI DeveloperInformation Developer,Data Scientist,AI DeveloperIBM Skills NetworkIB
2、M Skills NetworkStudent Dev Day:Agentic RAG:Autonomous Intelligence Meets Knowledge Retrieval(sprinkled with a hint of MCP)Agenda0102030405App overviewKey concepts overview+demoHands-On LabTakeawaysQ&AIBM TechXchange|2025 IBM CorporationBefore we start IBM TechXchange|2025 IBM Corporation3The presen
3、tation provides a detailed technical explanation,followed by a 15-minute hands-on lab session to apply what youve learned.0101Feel free to take your time with the lab.If you dont finish during this session,there are instructions at the end of the session to pick up where you left off.0202If you have
4、 questions,dont hesitate to ask during the hands-on lab session!0303Lets begin!What you will build by the end of this sessionIBM TechXchange|2025 IBM Corporation4What you will learn in this session0102030405060708Agentic AI&Multi-agent systemAgentsTasksTools and Tool CallingLangGraphAgentic RAGDocli
5、ngBonus:Model Context Protocol5What is agentic AI?An individual system or An individual system or program program that operates autonomously with agency.It can perform taskstasks,make decisionsmake decisions,and interactinteract with the environment or other systems independently.One AI system One A
6、I system to solve problems or achieve goals on behalf of a user or another system.Often relies on advanced AI models(LLMs,etc.)IBM TechXchange|2025 IBM Corporation6Conventional structure of an AI agentIBM TechXchange|2025 IBM Corporation7What is a multi-agent system(MAS)?A system where multiple auto