Build with AI 2026 | Vibe Coding, Agents & the Future of Development

Build with AI is a hands-on technical series designed to explore the frontier of modern software development. This year, we are diving deep into "Vibe Coding" , an emerging paradigm where natural language becomes the primary syntax, enabling developers to generate code, craft interfaces, and ship applications faster than ever before.

Through this online session of the 2026 series, participants will explore the evolution from traditional engineering workflows to AI-native, agent-driven development. Attendees will gain direct exposure to Google's rapidly evolving AI ecosystem and learn how modern developer tools are reshaping the future of building software.

What You'll Explore

• The Gemini Era | using Gemini models and LLMs beyond simple chat experiences

• Rapid prototyping workflows with Google AI Studio and modern AI tooling

• The new developer stack featuring Gemini CLI, ADK (Agent Development Kit), and agentic workflows

• The rise of AI Agents and how they are transforming software development lifecycles

Whether you are a student, mobile developer, cloud engineer, or AI enthusiast, this event is designed to help you move from simply using AI to confidently building with AI.

Session 1

Building High Performance Data Layers for Agents with Gemini Embeddings 2.0

Speaker

Haren Bhandari
Senior Developer Relations Engineer, Google, Tokyo

Session Overview

In this session, Haren will take a deep dive into Gemini Embeddings 2.0 and demonstrate how it enables high-performance AI Agent development.

Google recently introduced a natively multimodal Gemini Embeddings model capable of handling audio, video, images, and text with high precision. This unlocks powerful new possibilities for building advanced retrieval systems that provide AI Agents with richer context and significantly improved performance.

What You'll Learn

• The latest developments in AI and Agent ecosystems at Google

• Fundamentals of Agents, RAG (Retrieval-Augmented Generation), and Embeddings

• Different embedding models and the advancements introduced in Gemini Embeddings 2.0

• Real-world systems and architectures that can be built using multimodal embeddings

Session 2

Trace, Monitor and Evaluate your AI Agent

Speaker

Tarun Jain
Research Engineer at Kaivid Labs and Google Developer Expert (AI)

Session Overview

In this hands-on workshop, Tarun will guide participants through building a search engine tool using Qdrant and integrating it with a Google ADK AI Agent.

The workshop focuses on understanding the importance of tracing, monitoring, and evaluating LLM-powered applications in production environments. Participants will also explore Comet Opik, an open-source observability framework designed to trace and monitor AI Agents effectively.

What You'll Learn

• How to build and integrate search tools with AI Agents

• Using Qdrant for vector search and retrieval workflows

• Monitoring and tracing LLM applications in real-world systems

• Evaluating AI Agent performance using observability frameworks like Comet Opik

Get ready to transform your ideas into intelligent, production-ready AI applications.