Join us for a watch party exploring how to build a real-time multilingual broadcast application using Gemini 3.5 Live Translate, LiveKit, and Google Cloud Run.

In this technical walkthrough, Google DeepMind’s Thor Schaeff demonstrates an open-source application that translates live speech across languages in real time. We’ll follow the complete development flow, including setting up translation sessions, handling WebSocket connections, integrating LiveKit, deploying the application with Cloud Run, and exploring the codebase with Antigravity.

This session is useful for developers, students, and anyone interested in real-time AI, speech translation, cloud deployment, WebSockets, and Gemini-powered applications.

What we’ll cover:

• Real-time multilingual speech translation

• Gemini 3.5 Live Translate

• LiveKit integration

• Translation session management

• WebSocket connections and streaming

• Deploying with Google Cloud Run

• Exploring and understanding the open-source codebase

• Using Antigravity during development

Resources
Code: https://goo.gle/4z9lyPr
Documentation: https://goo.gle/4zgpuOt