| 2:00 AM | Registration and Pre-Event Networking | |
| 5:00 AM | Opening Remarks | |
| 5:20 AM | "The Agentic Developer: Coding Beyond Autocomplete" by Louie Miguel | AI coding tools are evolving beyond autocomplete into powerful coding agents that can understand context, use tools, and work across complex software projects. Developers can leverage Antigravity alongside Agent Skills, MCP, and Open Knowledge Format (OKF) to give AI agents the context, capabilities, and knowledge they need to plan, implement, test, and iterate on real-world software. |
| 6:05 AM | "Powering Generative UI in the Cloud" by Joshua de Guzman | Most AI features still answer in text, but an agent can answer with an interface instead. A2UI is Google's open protocol for generative UI: the agent describes the screen as JSON, and your app renders it with its own native widgets from a catalog you control. This session shows how to serve that UI from Google Cloud. We start with A2UI over MCP, where an MCP server on Cloud Run returns UI instead of raw data, so one server can drive Flutter, web and Angular clients. Then we move to A2UI over A2A, where an ADK agent on Gemini Enterprise Agent Platform (formerly Vertex AI) composes a full screen from the conversation and live data pulled through the BigQuery MCP server. We also cover IAM and agent identity so the agent only reaches what it should, and how to choose between templated and fully generative UI. You leave with a working architecture, a Flutter demo and a clear rule for picking the right transport. |
| 6:50 AM | "Trust, But Review: What Your AI Agent Decides Without Asking" by Miguel Kua | Agentic coding makes shipping faster, but it also makes edge cases easier to overlook. Miguel shows how Jules turns a short task into convincing code, where the hidden decisions are, and how to decide what to trust and what to review, even under a tight deadline. |
| 6:55 AM | "The Agile Mindset: Elevating Business Analysis in Modern Enterprise Delivery" by Van Orpilla | In modern enterprise engineering, technical complexity is a given, yet projects often stumble not from code failures, but from the translation gap between business strategy and technical execution. This talk explores how adopting an Agile mindset and elevating the strategic role of Business Analysis (BA) bridges that gap. Designed for tech industry professionals and leaders, we will examine how moving away from rigid requirements and toward collaborative value facilitation drives enterprise maturity, reduces waste, and aligns technical delivery directly with measurable business outcomes |
| 7:00 AM | Afternoon Tea Break | |
| 7:30 AM | "From Localhost to Zero Trust: Securing ADK Agents on Cloud Run with IAP" by Renzi Vidal | Running an AI agent on localhost:8080 is simple, but how do you share it with internal teams without exposing sensitive data or provisioning complex networking? In this 5-minute lightning talk, discover how to deploy a Google Agent Development Kit (ADK) agent to Cloud Run and lock it down using direct Identity-Aware Proxy (IAP) integration—no external load balancer required. We’ll cover how to containerize the agent, deploy with the --iap flag, and manage team access via Google Identity and IAM. Attendees will leave with a zero-cost, zero-trust blueprint to move agents from local terminals to production-ready corporate tools in minutes. |
| 7:35 AM | "From Raw Data to Charts in Google Colab" by Romark Cacho | Discover how to turn raw data into useful charts using Google Colab and Python. This lightning talk will show a simple process for loading data, exploring patterns, creating visualizations, and sharing results. Participants will see how Colab can support data analysis, classroom activities, research, and AI-assisted visualization. |
| 7:45 AM | "Testing AI-Powered Flutter Apps with Firebase AI Logic" by Cleo Credo | AI features make apps more capable, but they're hard to test. Model responses are non-deterministic, network calls are slow, and every test run can cost money. This talk shows how to unit test Flutter apps that use Gemini through Firebase AI Logic (formerly Vertex AI in Firebase).
You'll learn how to structure your code so AI calls sit behind testable interfaces, how to mock model responses for fast and repeatable tests, and how to cover edge cases like malformed output, empty responses, and API failures. We'll finish with practical patterns you can add to your own Flutter projects so AI features ship with the same confidence as the rest of your app. |
| 8:30 AM | "MCP, You Barely Know Her!" by Weiyuan Liu | Model Context Protocol (MCP) is quickly becoming the standard bridge between AI agents and the tools they need — but getting the most out of it takes more than a simple client-server setup. This session walks through MCP's real-world range: from spinning up a local client-server pair, to integrating remote services like Supabase and Google Maps, to building custom servers with FastMCP and MCP Toolbox. We'll also touch on production-ready deployment patterns and new agent workflow capabilities in ADK 2.0 — Graphing Flows and Human-in-the-Loop — plus a fun closing demo you won't want to miss. |
| 9:30 AM | Closing | |
