Frontend and backend code for Google Cloud’s “AI Demos” — visual, interactive examples of what the platform’s AI offerings can do. This is my fork of the official GoogleCloudPlatform/google-cloud-ai-demos repo, kept in my account for demo work and study; as a Customer Engineer specializing in AI on Google Cloud, having the demo source close at hand is genuinely useful.
The concepts
These are explicitly not production apps — they exist to show, not just tell. The repo covers multiple demos across products and use cases, including:
- Matching Engine — a text and image search engine powered by Vertex AI Matching Engine (vector similarity search).
- Time-series forecasting — a forecasting app built with React, Material UI, Cloud Run, BigQuery, and Vertex AI.
The hosted versions live at ai-demos.dev, so you can try them without deploying anything.
How to use it
The most common way I use this repo is as reference code: when a customer asks “what does a real app on Matching Engine look like?”, the answer is a working React frontend and a backend wired to actual Vertex AI APIs, not a slide. For deployment details, each demo’s directory in the upstream repo documents its own setup.
Language: TypeScript
Source code: github.com/avnit/google-cloud-ai-demos Upstream: github.com/GoogleCloudPlatform/google-cloud-ai-demos