AI Live + Labs Berlin

Today's organizations operate within a maze of disconnected applications and disconnected data, where critical insights are hard to find and people spend way too much time context switching. The first wave of AI gave us chatbot assistants, but these tools, while important, are often also app-bound and can end up reinforcing the silos. True transformation comes from agentic AI. which presents an opportunity to move beyond incremental gains and generate real value. Join us to discover how to build the next competitive advantage. 

Hear how customers are exploring and applying Gemini capabilities in their organizations. The discussion will cover the business priorities shaping their approach, practical applications, adoption considerations and early lessons from putting AI to work.

Imagine a world where your most complex tasks are simplified, your strategic thinking is amplified, and you are always one step ahead. This isn't a distant future; it's the new reality we want to build with you in our AI Workshop. In this 75 minute session, you will get hands-on experience bridging the gap between your daily executive functions and a new, AI-powered approach that will redefine your leadership.

How do you go from a business use case to a working AI agent in minutes? In this hands-on session, we explore the core foundations of building autonomous agents using the latest Gemini models, Google’s Agent Development Kit (ADK), and AI-assisted development workflows (agents-cli and coding agent skills). You will scaffold an agent from scratch, equip it with custom tools and structured workflows, and interactively test and iterate on its behavior in a live local playground.

Building a local prototype is the easy part; turning it into a reliable production system at company scale is where the real engineering begins. In this hands-on session, you will take an agent from local development to a deployed cloud solution—adding automated quality evaluation, production infrastructure, and end-to-end observability. We will also show how organizations can standardize and scale agent development across teams using extensible CLI workflows and customizable plugins tailored to their own infrastructure (such as Kubernetes, Cloud Run, or managed runtimes).

Unlock the full potential of Agent Development Kit (ADK) in this hands-on session. We begin by building robust agents capable of tool use and structured output, tested across CLI and Browser environments. We’ll then pivot to advanced system design, focusing on the Agent-to-Agent (A2A) protocol to connect and manage distributed multi-agent collaborations. Finally, you will learn how to transition your architecture from development to production on Google Cloud Platform, deploying your newly built agents using Cloud Run and the Google Agent Engine.

Today, the most successful developers aren't just writing code, they are orchestrating AI tools and agents to build better and faster. This session dives into this modern "Builder Stack”. You will learn to build with a variety of tools in the Google ecosystem - from idea, to generating a beautiful UI, building AI features, and integrating a secure, scalable backend.

 In this lab, you transform a standard BigQuery dataset into a governed data product and deploy an AI agent grounded in semantic business definitions. You will implement robust platform-level guardrails—preventing data mutations and masking sensitive columns directly through IAM, Data Catalog, and tool definitions rather than prompt engineering. You will configure and deploy the end-to-end solution using Antigravity.

In this lab, you expand the solution towards a zero-ETL architecture spanning AlloyDB and Spanner. BigQuery queries both operational databases in real time without pipelines or data staleness. A unified property graph connects all three systems, while an IAM-governed router directs queries to authorized agents. As in the previous lab, your main tools will be Antigravity and prompting.