About Vibe Commerce

This is a demo e-commerce site developed completely by vibe-coding with Gemini Code Assist.

Purpose

Vibe Commerce is a fictional e-commerce site built for the express purpose of demonstrating the powerful capabilities of Google Cloud's Vertex AI Search for Commerce. It's designed to illustrate how the AI capabilities of Vertex AI Search for Commerce enable key differentiators like industry-leading site search, personalized recommendations, and conversational commerce.

Key Features

This demo highlights several core features of Vertex AI Search for Commerce:

Site Search: Experience fast, relevant, and intuitive search results that understand user intent, not just keywords.
Product Recommendations: Discover products tailored to your browsing history and preferences with our "Recommended For You" carousels.
Conversational Commerce: Engage with our AI-powered chat assistant to find products, ask questions, and receive guidance through natural language conversation.

Technology Stack

This project is built on a modern, scalable stack of Google Cloud services:

	Vertex AI Search for Commerce	The core engine for search, recommendations, and conversational AI.
	Vertex AI Discovery Engine	Provides conversational answers to non-product search intents.
	BigQuery	Used for ingesting and storing the product catalog and user event data.
	Google Cloud Storage	Used for storing static assets like product images and icons.
	Cloud Run	Provides the serverless, containerized environment for hosting the Flask web application.
	Vertex AI Colab Enterprise	Used for data preparation and interacting with Google Cloud APIs in a notebook environment. The product catalog and product images were generated in Colab Enterprise using the Gemini and Imagen models.
	Gemini Model	Powers the generative AI capabilities, including the AI chat assistant. Primary model for generating the product catalog.
	Imagen Model	Google's text-to-image model, used for generating product images.

System Architecture

This diagram illustrates how the different Google Cloud services work together to power the Vibe Commerce application, from offline data preparation to online user interaction.

Inspiration

The design and spirit of this demo site are affectionately modeled after Toys "R" Us, a beloved childhood toy store that was a cultural icon in the 1980s. For many, it was a magical place of discovery and joy. Though the original stores have since closed, this project aims to recapture a small piece of that nostalgic experience within a modern, AI-powered e-commerce framework.

Author

This project was created by Jake Holmquist, Field CTO, Google Cloud at Valtech.

Source Code

The complete source code for this project is available on GitHub. You can explore the code, see how the features were implemented, and even contribute to the project here: https://github.com/cloud-jake/vibe-commerce.