Build with AI is a series of community-led technical workshops hosted by Google Developer Groups (GDG), designed to help developers explore practical ways to integrate artificial intelligence into real-world applications. These hands-on sessions focus on designing, building, and evaluating multi-agent systems, where multiple AI agents collaborate to solve complex problems.

Participants will gain experience working with agent-based architectures, experiment with coordination and communication strategies, and apply modern AI techniques to build intelligent, scalable systems. Through guided codelabs and practical exercises, attendees will not only build working systems but also learn how to measure and improve their performance.

Whether you’re a beginner exploring AI systems or an experienced developer looking to deepen your understanding of agent-based design, this series provides practical skills and hands-on experience that can be applied immediately.

Workshop Sessions:

1. Building a Multi-Agent System - Led by Andrew Sithole and Sylvia Dieckmann

(Codelab link 🔗)

Overview:

In this hands-on session, participants will learn how to design and implement a system where multiple AI agents interact to solve complex problems. The workshop introduces core concepts such as agent communication, coordination strategies, and system design.

Participants will:

• Understand the fundamentals of multi-agent systems.

• Learn how agents communicate and collaborate.

• Build a functional multi-agent system through guided coding exercises.

• This session provides a strong foundation for working with distributed AI systems and agent-based architectures.

2. From "vibe checks" to data-driven agent evaluation—Led by Thamu Mnyulwa and Ulrich Kotze

(Codelab link 🔗)

Overview:

This session focuses on evaluating the performance and effectiveness of multi-agent systems. Participants will explore key evaluation metrics, scenario-based testing, and methods for analyzing system behavior across different scenarios.

Participants will:

• Learn how to measure performance in multi-agent environments.

• Apply scenario-based testing techniques.

• Identify ways to improve efficiency, robustness, and coordination.

• By the end of this session, participants will have the tools to critically assess and refine agent-based solutions.

Prerequisites:

To ensure you can fully participate, please have the following:

• A laptop.

• Personal Google Account.

• Cloud Credits will be provided.

Important Notes:

• Bring a laptop with internet access and your cloud credits ready.

• The workshop is hands-on, so active participation is encouraged.

• Be prepared to collaborate, experiment, and engage with peers and facilitators.