| Arrival and Registration | Show ticket at the registration counter to check in and claim swags (if purchased). |
| [Hall 1] Opening Ceremony | Introduction to the event and NFC card tutorial. |
| [Hall 1 - Talk] Agentic AI with Gemini Enterprise by Wei Chung Low | (Intermediate)
Watch how Gemini Enterprise evolves from an AI assistant into a proactive business agent that understands the context of your work. We'll explore how it connects to your unique data sources to execute multi-step tasks, allowing you to spend less time searching for information and more time acting on it. |
| [Hall 1 - Talk] Supercharging Telco Tech Teams with AI by CelcomDigi | From smart testing to agentic AI engineering, Rachelle Foong and Stephanie Leong from CelcomDigi will take the stage and reveal one of Malaysia's connectivity giants' engineering playbook. |
| [Hall 2 - Workshop] Building AI Agents with ADK: The Foundation by Kamolphan Liwprasert | (Beginner)
This codelab, Building AI Agents with ADK: The Foundation, is the first part of a series designed to help you create your own intelligent AI agent using Google's Agent Development Kit (ADK). It guides you through the essential first steps, including setting up your Google Cloud environment, configuring a Python virtual environment, and crafting a simple, foundational conversational agent.
In this session, you will build a "Personal Assistant" agent powered by the `gemini-2.5-flash` model on the Vertex AI backend. You will learn how to define the agent's core logic, run it directly in the terminal, and interact with it through the ADK's development web UI for a richer chat experience, establishing a solid foundation for building more complex agents. |
| [Hall 1 - Talk] Fine-tune and Conquer: customizing Gemma for customer support magic by Matthias Baetens | (Beginner)
Gemma’s open models are powerful generalists — but to truly excel in specialized tasks, fine-tuning is the secret sauce. In this session, we’ll explore why and how to fine-tune Gemma, from data preparation to training strategy and evaluation. Using a real customer support use case, we’ll fine-tune Gemma to classify incoming requests, turning a generic model into a task-specific assistant. Along the way, we’ll cover parameter-efficient fine-tuning (PEFT), continued pre-training for domain adaptation, and how to operationalize your fine-tuned model on Google Cloud. Whether you’re new to fine-tuning or looking to streamline your workflows, this talk will give you the practical playbook to get started. |
| Lunch Break | Lunch is provided to everyone. Enjoy! |
| [Hall 1 - Talk] Agent2Agent (A2A) : A Beginner's Guide by Kuan Hoong Poo | (Beginner)
Agent2Agent (A2A): A Beginner’s Guide introduces newcomers to the rapidly growing world of autonomous agents and their ability to collaborate, communicate, and solve problems together. This session breaks down the fundamentals of agent-to-agent interactions, from core concepts and underlying architectures to real-world use cases where agents negotiate, share knowledge, and coordinate tasks. Designed for beginners, it will demystify technical jargon, highlight key challenges and opportunities, and provide a practical roadmap for understanding how A2A systems are shaping the future of AI applications. |
| [Hall 2 - Workshop] Practical AI for Web Developers by Tamas Piros | (Beginner)
As a web developer with over two decades in the industry, I've always been curious about the next big thing shaping our field. Recently, I took some time to dive into the rapidly growing world of AI - unpacking buzzwords like multimodality, agents, RAG, large language models (LLMs), tokens, and embeddings. I wanted to go beyond the hype and truly understand how these tools can enhance the way we build.
Now, I’m excited to share what I’ve learned in a practical, developer-friendly way. My hands-on workshop is designed to empower fellow web developers - whether you’re just starting your journey or have years of experience. To keep things simple and focused, we’ll stick to Node.js and the CLI (command line). Together, we’ll explore how to integrate powerful AI features directly into your projects.
My goal is to make AI approachable, translating complex concepts into something relatable and actionable - developer to developer. By the end, you’ll be equipped to create intelligent, cutting-edge web applications that stand out in today’s tech landscape.
|
| [Hall 1 - Talk] Build a Transformer LLM with JAX from Scratch by Lai Fong Leong | (Intermediate, Advanced)
JAX is a Python library for accelerator-oriented array computation and program transformation, designed for high-performance numerical computing and large-scale machine learning. As JAX is growing in popularity, let's explore how to develop a Transformer model with JAX.
In this session, you we will first understand the Transformer model architecture. Next, you will learn developing all the components required for a large language model (LLM) from scratch using JAX. |
| [Hall 1 - Talk] What Chrome's Built-In AI (Gemini Nano) Can Do and the Future of AI on the Browser by Warat Wongmaneekit | (Beginner)
On-device AI is revolutionizing web-based AI applications by leveraging the processing power of user devices instead of relying solely on cloud infrastructure. This approach enables faster, more responsive AI experiences with enhanced privacy. This session will guide developers through the latest trends in On-device AI, highlighting its potential and providing practical insights on how to seamlessly integrate it into your applications. Also, this session will provide a demo of Gemini in Chrome to make developer able to understand what will happen in the future. |
| [Hall 2- Workshop] Speed Building ADK Voice Agents: A Workshop of Tips, Tricks and Hacks to Build Winning Agents by Sam Witteveen | (Intermediate)
Step into the world of rapid prototyping and learn what it takes to build a voice agent in just an hour. This workshop will equip you with the essential tips, tricks, and mindset for high-speed agent development.
Drawing from the experience of winning the ADK (Agent Development Kit) Bake Off Hackathon in San Francisco, Sam will guide you through the process of creating a powerful voice agent under pressure.
What You Will Learn:
(a) A Plan for Success: How to create a well-defined vision for your agent, a modular approach, and the structure for a successful project.
(b) Rapid Prototyping Techniques: Learn how to use context stuffing of Gemini to quickly bootstrap your agent with relevant information, prompts and skills.
(c) How to Leverage Pre-Built Tools: See how to integrate ADK Web with MCPs and other tools to create a functional and engaging user experience.
(d) How to Overcome Challenges: Gain insights into handling unexpected issues with frameworks, unexpected LLM responses and unplanned user behavior.
(e) Live Demo and Q&A: Watch a live demonstration and get your questions answered.
Whether you're a seasoned agent developer or just starting, this workshop will provide you with tools and inspiration to build innovative voice agents quickly and efficiently. |
| Teabreak | Short break with light refreshments at the open area. |
| [Hall 1 - Talk] Unleashing Gemini superpower with Chrome Extension by Vin Lim | (Beginner, Intermediate)
Explore the incredible potential in the combination of Gemini and Chrome extensions. Chrome Extensions can help you create tailored web experiences, including for video and audio. With Gemini's powerful multimodal APIs, you can build completely new experiences. |
| [Hall 2 - Workshop] Agentic AI with Google's Agent Development Kit and Firebase by Cleo Credo | (Intermediate)
AI is moving beyond simple chatbots into agentic AI. Agents that can reason, plan, and take action on behalf of users. Rather than simply responding to queries, these agents can proactively execute tasks, connect with tools, and adapt based on context.
This workshop will explore how to build intelligent agents using Google’s Agent Development Kit (ADK) together with Firebase. It will highlight how Firebase provides the foundation for memory, persistence, and real-time interactions that enable agents to deliver real value.
By the end of the session, attendees will learn first-hand how Firebase can serve not just as a backend, but as the knowledge base, identity layer, and action hub that powers agentic AI. |
| [Hall 1 - Workshop] Social Media Product Video Generation with Nano Bana and Veo with AI Studio by Shang Yi Lim | (Beginner)
Learn how to use AI Studio, Vertex AI Studio, Nano Banana and Veo to create and deploy a product video generator for social media use. |
| [Hall 2 - Workshop] Leveraging AlloyDB and Vertex AI for an LLM and RAG-Based Chat Application by Ananda Dwi Rahmawati | (Intermediate)
In this session, you'll learn how to build a reliable and scalable infrastructure for Retrieval-Augmented Generation (RAG)-enabled generative AI applications using Google Cloud’s Vertex AI. RAG enhances the performance of generative AI by integrating context-specific, non-public data into prompts to improve the accuracy of Large Language Model (LLM) responses. You'll explore how to leverage Google Cloud's robust tools, such as AlloyDB for vector storage and semantic search, and how to use Vertex AI’s powerful capabilities, including the Gemini Pro multimodal model, to create an intelligent, scalable solution. This session will cover the architectural considerations, best practices, and key integrations to deliver an efficient, production-grade RAG-powered AI application. |
| Closing Ceremony | Wrap up the day with a lucky draw! |
