Running GenAI Apps on Google Cloud

Deploy Generative AI Apps with Cloud Run in Google Cloud with confidence and speed: In this session, you will learn about Gen AI use cases in Cloud Run and how to combine LangChain on Cloud Run with Cloud SQL’s pgvector and Vertex Endpoints to create generative AI applications.

In the presentation, Steren Giannini and Lisa Shen from Google Cloud discuss deploying Generative AI (GenAI) applications using Cloud Run on Google Cloud. They highlight how GenAI is transforming interactions with technology, both in AI-assisted development and in applications that incorporate language and other AI models.

The presenters introduced GenAI apps, emphasizing their transformative role in technology interaction. They specifically mentioned LangChain, a popular framework for building language-based applications, which integrates seamlessly with Cloud Run for easy and fast deployment. They also discussed the architecture of GenAI apps, focusing on a model called Retrieval-Augmented Generation (RAG). This model enriches language model responses by querying a database with context data, improving accuracy.

Lisa Shen demonstrated the deployment process and operation of a GenAI app using Cloud Run, showcasing the ease of integrating with other Google Cloud services like Cloud SQL and Vertex AI. The demo involved setting up a web app to answer queries about Cloud Run features using data indexed from release notes stored in a vector database. They emphasized Cloud Run’s scalability, cost-effectiveness (scaling down to zero when not in use), and ease of use, which simplifies the deployment and management of containerized applications without the need to manage infrastructure.

Colab for the demo: https://codelabs.developers.google.com/codelabs/build-and-deploy-a-langchain-app-on-cloud-run