Speakers
- Sharmila Devi (AI Consultant, Google)
- Gopala Dhar (AI Engineer, Google)
Abstract
This tutorial offers a practical deep dive into building and deploying the next generation of AI applications by mastering the end-to-end design of multi-agent systems. As the industry moves beyond monolithic models, this session equips AI researchers, developers, and practitioners with the critical skills needed to architect sophisticated agentic solutions. Using the Agent Development Kit (ADK), attendees will engage in interactive labs to build agents from the ground up, empower them with external APIs, and deploy them to a scalable Agent Engine. Participants will leave with the demonstrated ability to design, implement, and manage complex multi-agent systems ready for real-world challenges.
Target Audience
- Academic researchers
- Industry AI developers
- AI Leaders
- Graduate students in AI and computer science
- Agentic AI Practitioners
- Software Developers
Outline and Description of the Tutorial
This tutorial offers developers and engineers a practical, hands-on introduction to building, empowering, and deploying sophisticated agentic systems using the Agent Development Kit (ADK). Through a series of focused modules and interactive labs, participants will progress from fundamental concepts to deploying a complete multi-agent system. This session is designed to equip attendees with the skills needed to leverage the full potential of the ADK and Agent Engine in real-world applications.
Key Topics Covered:
- Getting Started with the Agent Development Kit (ADK): A foundational overview of the ADK architecture, core components, and setup process for a local development environment.
- Building Multi-Agent Systems: Practical guidance on designing and implementing systems where multiple agents collaborate to accomplish complex tasks, including communication patterns and state management.
- Empowering ADK Agents with Tools: Techniques for extending agent capabilities by integrating them with external APIs, databases, and other services through the use of tools.
- Deploying ADK Agents to Agent Engine: A step-by-step guide to packaging, deploying, and managing agents on the Agent Engine for scalable and robust operation.
By the end of this tutorial, participants will be able to:
- Define the core architecture and principles of the Agent Development Kit.
- Build and run a standalone agent on their local machine.
- Design a multi-agent system where agents interact and share information effectively.
- Integrate custom and third-party tools to enhance an agent’s functionality.
- Deploy, monitor, and manage agents.
Reading List
For a First Impression (Optional Pre-reading): These resources provide a high-level overview of the multi-agent field and its modern context, perfect for someone looking to understand the significance of the topic before diving in.
- Book: An Introduction to Multiagent Systems by Michael Wooldridge
- Survey Paper: A Survey of Large Language Model based Autonomous Agents (Lei Wang, Chen Ma, Xueyang Feng et al., 2023 on arXiv)
Expected Prerequisite Knowledge (Please Read Before Attending): To ensure everyone can keep up with the hands-on labs, attendees should have a working understanding of the concepts in these key papers.
- Foundational Paper on Tool Use: “ReAct: Synergizing Reasoning and Acting in Language Models” (Shunyu Yao, et al., 2022).
- Foundational Paper on Multi-Agent Collaboration: “Generative Agents: Interactive Simulacra of Human Behavior” (Joon Sung Park, et al., 2023).
For Further Information (Post-Tutorial Exploration): For those who want to dive deeper into the theory or explore advanced research topics after completing the tutorial.
- Advanced Book: Multiagent Systems: Algorithmic, Game-Theoretic, and Logical Foundations by Yoav Shoham and Kevin Leyton-Brown.
- Advanced Topic – Reinforcement Learning: Multi-Agent Reinforcement Learning: Foundations and Modern Approaches by Stefano V. Albrecht, et al.
Vertical
Generative AI Models, AI in Education and Agentic AI
Timeline
2 hours