Yes. If your boards publish volume (in liters), length, and recommended ability, the bot can match a surfer's weight and skill to in-stock options. A 75kg intermediate gets a 38 to 44 L range with two or three in-stock candidates. For shops that maintain a volume calculator on the site, the bot quotes the calculator's logic faithfully rather than inventing alternate guidance. Custom fields for rocker profile, rail shape, and tail can also feed the bot's recommendations.

If your boards tag fin setup (single, twin, thruster, quad, 2+1, five-fin) and you carry fin compatibility data, the bot can match fins to boards in stock. A customer asking "do you have FCS II thrusters for this board" gets a yes/no plus in-stock fin sets. For shops carrying multiple fin systems (FCS, FCS II, Futures), the bot quotes compatibility per board, which is a common source of confusion for new customers.

If you publish a water temp range for your region and a wetsuit thickness guide, the bot quotes them. A customer asking about May conditions in a 14-16 C region gets a 3/2 to 4/3 recommendation with a note about hooded vests for longer sessions. The bot stays grounded in your published guide and does not invent thicknesses for conditions you have not described. Care content (rinsing, drying, neoprene cement repair) lives in the vector store for follow-up questions.

If your demo fleet and rentals are managed through WooCommerce Bookings or similar, the bot can read availability for the dates the customer requests. For shops that manage rentals at the counter, the bot quotes the demo rate sheet and rental policy from the policy page. The bot describes the booking flow and either points to the booking page or recommends calling the shop directly during business hours.

Yes, if your ding repair policy is published. The bot quotes typical turnaround ("standard ding repair 5 to 7 business days, rush 48 hours for an additional 20 EUR") and walks the customer through the drop-off process. For complex repairs (snapped board, deep delamination), the bot describes that an estimate is required after the shaper inspects and routes the customer to drop the board off rather than promising specific repair pricing.

If your catalog tags boards by use case (everyday shortboard, fish, fun-shape, mid-length, longboard, gun, step-up), the bot can match a customer's described conditions to in-stock candidates. A surfer asking for a board for a Mentawai trip gets recommendations grounded in step-up and travel-friendly tags. The bot stays factual: it surfaces options rather than pushing a single shape, and it references the shaper's notes if those are published on the product page.

In your WordPress database. Each message is logged with model name, token usage, and page URL. Nothing routes through Sleek infrastructure: the data path is browser to WordPress to your model provider on your API key. For shops with EU customers, this keeps the data-residency story simple. Conversation logs are also useful for spotting which products generate the most questions, which feeds product copy improvements.

Multibot lets you run several bots with different prompts and display rules. A shortboard bot can lean into volume per kg, rocker profile, and performance shapes; a longboard bot can be friendlier and focus on stability and noseriding glide; a wetsuit bot can lean into thickness and care. Display conditions scope each bot to the right section, so the right specialist appears at the right moment without juggling multiple chatbot vendors.