One engine.
Every kind of retail.
Generic assistants fall short wherever the purchase takes expertise and the category has rules. shopsupport.ai builds agents trained on your catalog, your decision logic and your rules, so every recommendation is accurate, in stock, and compliant in the vertical it serves.
A personal shopper for every department
A retail catalog is wide: thousands of products across departments, each with its own vocabulary and its own rules. Shoppers describe a situation ("everything for a newborn's first bath", "a detergent for wool and delicates") and site search wants the product's name. The personal shopper hands the need to the right department expert, answers from what is actually in stock with the reason each pick fits, and builds the cart across departments when the need spans them.
Claims are governed per category. Cosmetics stay inside cosmetics claims rules, supplements use authorised health claim wording only, baby food is presented neutrally, biocides carry their mandatory safety sentence, and non prescription health products are described from the label, with anything beyond it referred to a pharmacist. Every release passes a claims test suite before it ships.
Ready for European stores: AI disclosure under the AI Act, data kept in the EU, no identity data collected, accessible to WCAG 2.1 AA. Several languages on one catalog, with the store's own product names, so every shopper sees the same products at the same prices.
Sell the pairing, not the shelf
Wine and spirits shoppers rarely search by producer. They describe a dinner, a gift, a mood or a price, and expect the shortlist a good floor sommelier would give: grape, region, style, why it fits. The agent turns that description into structured intent, filters live stock, and explains the pairing in one reply.
Compliance sits in the answer, not in the footer: age verification before any recommendation, shipping eligibility by destination state or country, quantity limits where they apply, and responsible drinking language. Allocations, club pricing and case discounts are catalog facts the agent reads, never promises it invents.
Taste is data. Producer notes, tasting descriptors, sweetness, body and tannin become filters, so "something like the Barolo I had, but cheaper" resolves to a Langhe Nebbiolo at a third of the price, in stock.
Sell the look, size it right
Apparel shoppers know the occasion and how they want to feel, not the product name. The agent turns "wedding guest in July, not too formal" into a complete look from live inventory, checks the size for each piece, and explains why the pieces work together.
Fit is the return problem, so the agent reads size charts, model measurements and what customers of a similar size kept, and says when to size up or down. Returns fall when the first order fits.
Merchandising rules are respected: new season first, markdowns when the shopper asks for a budget, brand exclusions and regional catalogs. Every look ends in an add to bag for the whole outfit or a saved shortlist.
Boutique care, at every hour
Jewelry and watches are bought for someone else, on a date that matters. The agent asks the two questions a good boutique asks, who is it for and what is the occasion, then works within the budget across the metals, stones and styles the shop actually has.
Materials are explained, not marketed: gold purity, plating, stone grading, movement type, water resistance. Certifications and provenance come from supplier data and are shown as such.
The practical parts close the sale: ring sizing guidance, engraving options, gift wrapping, delivery dates and the returns policy, all in the same conversation, with the shortlist saved for a partner to approve.
Turn every shopper into a guided buyer
Dispensary shoppers don't browse like typical ecommerce customers. They ask about effects, potency, consumption method, and what's actually in stock right now. shopsupport.ai agents connect to Dutchie and answer with live inventory, never recommending an out of stock strain.
Compliance is built in. Agents enforce daily purchase limits, age verification prompts, and state specific regulations before every recommendation. A shopper asking for "something strong" gets matched to products within their remaining daily allowance, not just the highest THC on the menu.
Strain discovery goes beyond category filters. Our agents understand indica vs. sativa effects, terpene profiles, and cross category intent, surfacing edibles, vapes, and flower that match what the customer is actually trying to accomplish.
Support That Closes The Ticket
The support agent reads the order, the return, the shipment and the account state through your systems, applies your policy, and takes the action the case needs: refund, replacement, address change, cancellation, subscription change. It runs on the same connection to your store as the shopping agent, in every industry above.
Generic AI vs. Domain Trained Agents
Your Store Deserves AI That Knows Your Catalog.
Tell us about your store and the system your catalog lives in. Every engagement starts with a scoped launch on your catalog and your market, then a platform fee plus usage.