AI Merchandising Engine

CS-Cart AI Merchandising Engine — Intelligent Personalized Recommendations

Increase conversions and maximize revenue with the powerful AI Recommender by Ecartify — now featuring smart recommendations across Products, Categories, Blog Posts, and Best Deals. This intelligent recommendation engine analyzes real shopper behavior — including searches, product views, cart actions, and completed orders — to automatically display the most relevant content at the perfect moment.

Powered by advanced vector search technology and AI embeddings, the addon understands product meaning, customer intent, and behavioral patterns — delivering truly personalized suggestions that feel natural and accurate across every part of your store.

Core AI Technologies

• Vector Database Powered by Qdrant: Uses high-performance vector similarity search to instantly match customers with the most relevant products, categories, blogs, and deals.
• AI Embeddings by Cohere: Converts product, category, and blog data into intelligent 1024-dimension embeddings that understand context and semantics — not just keywords.
• Multilingual Support: Embeddings are generated and stored per language, ensuring accurate recommendations for every storefront language.
• Real-Time Behavior Learning: Continuously learns from customer interactions to improve recommendation accuracy over time.

What Gets Recommended

• Products: Semantically similar products based on name, description, and meta keywords — embedded via Cohere and matched via Qdrant vector search.
• Categories: Related categories discovered by averaging product vectors per category — no manual tagging needed, fully automatic.
• Blog Posts: Relevant blog content matched to the visitor's current context — keeps shoppers engaged and on-site longer.
• Best Deals: Highlight your top offers intelligently based on behavioral signals and product relevance.

Smart Recommendation Capabilities

• Behavior-Based Personalization: Learns from Search, View, Add to Cart, and Order activity to personalize every suggestion.
• Weighted Activity Scoring: Adjust how much each customer action influences recommendations.
• Semantic Similarity Matching: Understands product and content meaning — not just category or tag matching.
• Flexible Display Blocks: Place recommendation blocks anywhere using the CS-Cart layout manager.
• Optional Add-to-Cart Button: Enable or disable direct cart actions inside recommendation cards.
• Fully Responsive Design: Optimized for mobile, tablet, and desktop storefronts.

Sync & Management Tools

• Paginated Cron Sync: Products and blogs embed in batches via cron — handles large catalogs without timeouts.
• Category Auto-Embedding: Category vectors are built automatically by averaging their product embeddings — no Cohere API calls needed for categories.
• Resync Tools: Instantly re-push all product, category, or blog vectors from your database to Qdrant after a server reset — free, fast, no API cost.
• Smart Skip Logic: Already-embedded content is skipped automatically on every run — safe to schedule without duplication.
• Per-Collection Management: Products, categories, and blogs are stored in separate Qdrant collections for clean, independent control.