In this talk, we will walk through a hands-on application that uses Google’s Vertex AI Platform and monitoring from DataDog to build a podcast search engine. We will start by downloading episodes from a podcast RSS feed, then use a speech-to-text model to create transcripts. Next, we will use Vertex AI’s Embedding API to create vector embeddings for each chunk of the transcript. These embeddings will be stored in a vector database, which allows us to perform semantic search across all of the conversations that have ever happened on the podcast. Finally, we will build a user interface that allows users to ask questions about the podcast in a chat-like manner. The Vertex Large Language Model will then answer these questions based on the context from the text embedding. This application demonstrates how Google’s Vertex AI Platform can be used to gain insights into your own data. By using DataDog to monitor the infrastructure, you can ensure that your application is running smoothly and efficiently. We will provide code samples so that you can easily set up and run this application yourself.