This session demonstrates how to use the Gemini API in Vertex AI to generate an engaging multi-speaker podcast using studio voices in the Text-to-Speech API.

This can be useful for creating interviews, interactive storytelling, video games, e-learning platforms, and accessibility solutions.

The steps performed include:

Load a PDF file from a Google Cloud Storage bucket or public URL
Summarize the content using Gemini 2.0 Flash
Return a pre-defined JSON schema using Controlled Generation
Create a multi speaker conversation from the JSON script using Text-to-Speech.
Generate the audio as MP3 file.