AI search for PrestaShop that navigates your entire catalog — combinations and all

PrestaShop is one of the most capable open-source ecommerce platforms available. Its product data model is remarkably sophisticated — products with combinations that each carry independent pricing, stock, and reference codes; deeply nested categories; manufacturers and suppliers as first-class entities; feature values and attributes that describe products with structured precision. For catalog management, this structure is a strength. For search, it becomes a liability. The default PrestaShop search indexes product names and descriptions, occasionally attributes if configured correctly, and returns a flat list of results ranked by keyword relevance. The entire combination system — the color/size/material matrix that defines how products actually exist in inventory — is invisible to the search engine. A shopper looking for "blue linen shirt in size 42" gets results for any shirt that mentions "blue" or "linen" or "42" anywhere in its text, regardless of whether that specific combination exists.

This gap between data richness and search capability is not unique to PrestaShop, but PrestaShop's combination system makes it particularly acute. On platforms without structured variants, the search problem is simpler: find products matching keywords. On PrestaShop, the real question is often more specific: does this product exist in this combination, at what price, and is it in stock? The default search cannot answer that question because it operates at the product level, not the combination level. Third-party search modules improve keyword matching — weighting attributes, boosting exact matches, adding faceted filtering — but they still fundamentally work by matching words to text fields and presenting results as a list. They cannot interpret "I need a waterproof hiking boot in brown, size 42, under 120 euros" as a structured query with product type, material attribute, color variant, size combination, and price constraint. That interpretation requires understanding language, not just matching it.

Asyntai's AI search for PrestaShop replaces keyword matching with conversational product discovery. Instead of a search box that returns a paginated list, shoppers interact with an AI chat widget that understands natural language, searches the live product catalog through a Real-Time Data Feed, and returns matching products as Dynamic Product Cards — visual cards with product images, prices, descriptions, and buy buttons displayed directly inside the chat. The conversation format means shoppers can refine, compare, and ask follow-up questions without restarting the search. "Show me hiking boots." "Something waterproof." "Do you have that in brown?" "What about a size 42?" Each message builds on the last, narrowing results conversationally — exactly how a shopper would interact with a knowledgeable sales associate in a physical store.

The foundation of this capability is the Real-Time Data Feed. PrestaShop's built-in Webservice API is one of the platform's most underutilized features. It exposes the entire catalog as structured data — products, combinations, categories, manufacturers, stock quantities, images, feature values, attributes — accessible via authenticated REST endpoints. You point Asyntai at your Webservice API URL (or a CSV product export, if you prefer), and the AI indexes your catalog. This is not a one-time import. The feed is live. When you change a price in the back office, the next shopper who asks about that product sees the updated price on the product card. When a combination goes out of stock, it stops appearing in results. New products added to your catalog are indexed within 24 hours. The AI's product knowledge is always synchronized with your actual inventory.

Dynamic Product Cards are what transform the chat from a text conversation into a visual shopping experience. When the AI finds products matching a shopper's request, it renders them as rich cards — each showing the product image, name, price (specific to the matching combination if applicable), a short description, and a "View Product" button linking directly to the product page. Multiple matches display as a horizontal carousel the shopper can swipe through. The cards appear inline in the conversation, surrounded by the AI's commentary — "The first two are waterproof-rated, the third is water-resistant but lighter" — creating a guided shopping experience that no search results page can replicate. Shoppers see products in context, with the AI acting as curator and advisor simultaneously.

The combination-awareness of this system deserves particular emphasis because it solves a problem that PrestaShop merchants know intimately. Consider a fashion store selling shirts in 5 sizes, 8 colors, and 3 materials — 120 combinations per product. A shopper searching for "white cotton shirt XL" on the default search might find the base product, but they have no idea whether the white/cotton/XL combination is actually in stock until they click through and manually select each attribute from dropdowns. With Asyntai's AI search, the answer is immediate. The AI checks the data feed for that specific combination, confirms availability, shows the correct price (which may differ by combination), and presents it as a product card. If white/cotton/XL is out of stock but white/cotton/L is available, the AI can say so. If the same shirt in linen costs more, the AI shows the linen price. This combination-level intelligence is native to how the AI processes product data — it doesn't require custom configuration or special attribute mapping.

Multilingual search is where PrestaShop stores gain an advantage they may not realize they need. PrestaShop's native multilingual support means many stores already maintain product data in multiple languages. But the default search typically operates within the current language context — a French-language visitor searching on the French version of the site searches French product names. What happens when a tourist visits your French store and searches in English? Or when an expat searches in their native language on a store configured for the local language? The default search fails silently, returning no results because the keywords don't match the current language's product names. Asyntai supports 36 languages and auto-detects the visitor's language. A German-speaking visitor on your French PrestaShop store can ask "wasserdichte Wanderschuhe" and the AI searches the same product data, understands the intent (waterproof hiking boots), finds matches, and responds in German — with product cards showing data from your catalog. The language barrier that would have been an invisible lost sale becomes a completed product discovery.

The conversational nature of the search solves iterative refinement in a way that no filter sidebar can match. On a traditional PrestaShop search results page, a shopper who wants to refine their results has to use faceted filters — clicking checkboxes for category, price range, attributes. This works reasonably well when the shopper knows the catalog's taxonomy. It fails when they don't. A shopper looking for "a gift for my wife who runs marathons, budget around 80 euros" has no filter combination that expresses this. They would need to navigate to the running category, guess which subcategory to choose, set a price filter, and hope something gift-worthy appears. In a conversation with Asyntai's AI, they simply state what they need. The AI searches across categories, understands that marathon running implies certain product types (performance shoes, technical apparel, running accessories), applies the budget constraint, and returns a curated selection as product cards. The shopper never needs to understand your category structure. The AI translates their intent into catalog queries.

Custom Tools extend the AI beyond product discovery into transactional territory. PrestaShop's Webservice API is bidirectional — it can read data and perform actions. By connecting Custom Tools, the AI can make live API calls during a conversation. A returning customer asks "where is my order?" — the AI calls your order endpoint, retrieves the status by reference number, and reports back. A shopper asks "is this jacket available in your Madrid store?" — the AI calls your stock endpoint filtered by warehouse. A shopper wants to know if a cart rule applies to a product — the AI checks. Each Custom Tool you connect adds another action the AI can perform without the shopper navigating away from the chat. The search conversation becomes a shopping conversation, handling the full journey from "I'm looking for..." through "is this in stock?" to "where's my order?" — all in the same chat window.

Installation on PrestaShop is deliberately simple. The widget is a single JavaScript snippet — a script tag that loads asynchronously. You can add it via a custom HTML module in PrestaShop's back office (any module that allows HTML in the header hook — "Custom text block" or similar), or paste it directly into your theme's header.tpl file. No module marketplace submission, no compatibility conflicts, no dependency on specific PrestaShop versions. The snippet works on PrestaShop 1.7, 8.x, and any theme because it loads independently of the page's DOM structure. It doesn't modify your templates, doesn't inject CSS that conflicts with your theme, and doesn't add database tables. One snippet, positioned in the header, and the AI search widget appears on every page of your store.

The economics of AI search on PrestaShop come into focus when you consider the cost of failed search. PrestaShop analytics typically show that shoppers who use the search function have higher purchase intent than browsers — they know what they want and are actively looking for it. When that search returns poor results or, worse, no results, you lose not a casual browser but a motivated buyer. These are your highest-intent visitors, and the default search fails them regularly. Every "waterproof jacket for autumn" that returns nothing, every "blue shirt size M" that returns all shirts regardless of color or size, every "gift under 50 euros" that the search engine cannot parse — each represents a buyer who tried to give you money and was turned away by your own search tool. AI search intercepts those queries and serves the products that should have appeared in the first place.

PrestaShop stores with large catalogs and many combinations benefit disproportionately from AI search. A store with 500 base products and an average of 20 combinations each has 10,000 purchasable items. The default search treats this as 500 products. The AI, connected to a data feed that includes combination data, treats it as 10,000 purchasable options, each with its own price and availability. When a shopper asks a specific question — "do you have this in red, size S?" — the AI gives a specific answer. This specificity directly reduces the click-through-to-abandonment pattern that plagues stores with complex variant structures, where shoppers find a product via search, click through, discover the combination they want is unavailable, and leave. The AI catches the mismatch before the shopper wastes a click.

Category and manufacturer navigation is another area where conversational search outperforms traditional navigation. PrestaShop stores with deep category trees — Electronics > Computers > Laptops > Gaming Laptops — require shoppers to know where in the tree their desired product lives. Many don't. A shopper looking for "a laptop good for video editing under 1500 euros" might navigate to Laptops but miss that the relevant products are split across "Professional Laptops" and "Content Creator Laptops" in your taxonomy. The AI doesn't care about category hierarchy. It searches the flat catalog, evaluates products against the shopper's stated needs, and returns the best matches regardless of which category they're filed under. Manufacturer-specific queries work the same way — "do you carry Samsung monitors?" triggers a search filtered by manufacturer, returning results organized by what the shopper asked for rather than by your internal category structure.

The gap between what shoppers expect from search and what PrestaShop delivers is growing. Visitors in 2026 have been conditioned by conversational AI — they've talked to ChatGPT, asked Siri complex questions, used Google's AI overviews. They expect search to understand them, not require them to guess the right keywords. A PrestaShop store with a keyword search bar is asking visitors to translate their needs into your catalog's vocabulary. Asyntai's AI search inverts that relationship. The shopper speaks naturally. The AI translates. The products appear as visual cards with correct prices and buy buttons. That's the interaction model shoppers already expect from the best online experiences — and it's now available for any PrestaShop store, regardless of size, theme, or technical complexity.

For PrestaShop agencies and freelancers who manage multiple client stores, AI search represents a high-value add-on. A PrestaShop store without conversational search is indistinguishable from every other store using the default search module. A store with AI-powered product discovery — where shoppers can describe what they want and see matching products appear as visual cards in a chat — immediately stands apart. The installation is per-store, per-widget, and takes minutes. The data feed connects to the existing Webservice API that every PrestaShop store already has built in. There's no module to maintain, no version-specific compatibility to track, and no theme modifications to manage. For agencies managing dozens of PrestaShop stores, it's a differentiation layer that scales without adding technical overhead to each deployment.

The stores that will gain the most from AI search share a common trait: they sell products that shoppers describe by need, not by name. Fashion ("something formal but not too stiff, navy, for a summer event"), sporting goods ("trail running shoes for rocky terrain, wide fit"), home furnishings ("a dining table that seats eight, extendable, modern style, under 800 euros"), specialty foods ("organic gluten-free pasta variety pack"), beauty and cosmetics ("anti-aging serum for combination skin"). In every case, the shopper has a clear mental picture but no product name. PrestaShop's default search requires product names or at least close keyword matches. Conversational AI search requires only a description of need — and it translates that description into the products sitting in your catalog, displayed as visual cards the shopper can browse, compare, and buy. That translation — from what the shopper wants to what you sell — is the fundamental value of AI search, and it works whether your catalog has 50 products or 50,000.