Yes. WooCommerce stores stock per flavour in postmeta. SleekAI reads it on every turn, so a 9 am Saturday visitor gets a real count for maple bacon, apple fritter, and the rest. When the count hits zero the next reply reflects it and suggests an in-stock alternative without making the customer refresh the menu page.

Map a rotation taxonomy where each flavour is tagged with the days it runs. The bot then knows maple bacon is Friday and Saturday only and matcha glaze is weekends. On a Tuesday it can apologise that the bourbon caramel is not on today and suggest a Tuesday lineup flavour instead.

Yes. Dozens and 24 packs are usually variable products with their own postmeta pricing. SleekAI reads the variation row directly, so a dozen mixed at $36 quotes as $36 with the $3 saving over single pricing called out. A 24 pack at $66 is quoted the same way without any arithmetic in the bot's head.

Yes, as long as the pre-order is placed before the cut-off. Map the cut-off time as a variable. The bot can confirm a Monday 7 am pickup for orders placed by 4 pm Sunday, and politely refuses orders that miss the cut-off. It can also suggest the next eligible pickup window.

Vegan and gluten-free donuts often have a daily cap. Map the cap and a daily counter in postmeta. The bot can confirm that today's vegan glazed cap of 6 has 4 left and refuses to commit more than the remaining quantity to a pre-order. That stops over promising on tight production days.

Yes. Display conditions support post type, product category, tag, user role, and URL pattern. A menu bot focuses on flavour stock and prices, while a pre-order bot specialises in quantity, mix, lead time, and pickup. Both read the same WooCommerce database with no duplication.

Map your delivery zones and lead times as variables. The bot can quote a $7 delivery fee to a nearby zip with a 30 minute window, plus a $25 minimum order. Outside the delivery zone it pivots to pickup and quotes the closest available slot from your booking plugin's database.

Yes. Each chat is stored with the model name, token usage, and the page URL it ran on. Repeated questions about a missing rotation tag or an unclear pre-order cut-off feed straight back to the rotation taxonomy and form copy, so the menu page and the bot improve in lock step over time.