About CGB Visual Search

We're building the future of e-commerce product discovery using artificial intelligence.

Our Mission

Traditional text-based search doesn't work when customers can't describe what they're looking for. CGB Visual Search bridges this gap by letting shoppers upload any image and instantly find matching products.

Built on OpenAI's CLIP model — the same AI technology used by the world's leading tech companies — our visual search engine understands images at a deep semantic level, not just pixel matching.

How It Works

When you sync your WooCommerce products, we generate a unique AI "fingerprint" (embedding) for each product image using the CLIP neural network. These embeddings are stored in a high-performance vector database (Qdrant).

When a customer uploads a search image, we generate its embedding and find the closest matching products using cosine similarity — all in milliseconds.

Technology Stack

🧠 CLIP AI Model

OpenAI's ViT-B-32 model for state-of-the-art image understanding and feature extraction.

⚡ Qdrant Vector DB

High-performance vector similarity search capable of querying millions of embeddings in milliseconds.

🚀 FastAPI Backend

Modern, high-performance Python web framework with automatic API documentation.

🔌 WooCommerce Plugin

Native WordPress plugin with one-click installation and automatic product syncing.