Yes. If you publish today's whole-fish, raw-bar, and oyster list on the menu page each afternoon, the bot reads it live. A guest asking at 19:00 about the catch gets tonight's actual list, not yesterday's. For restaurants that update the catch every service, this is the most useful capability on the page because the answer is always current.

Yes. If your oyster list is published with farm, bay, region, and flavour notes, the bot reads each oyster and matches the guest's preference. Mild, briny, sweet, copper-finish, the bot maps preference to the right oyster on tonight's list. It will not invent oysters you do not have, so the recommendation always corresponds to one of today's six or eight varieties on the board.

Line-caught, day-boat, MSC-certified, ASC-certified, and locally sourced tags all read from the menu post if you publish them. A guest asking 'is the tuna line-caught' gets your real sourcing answer. The bot will not invent a sustainability claim, so if you have not specified, it offers to put the guest in touch with the team rather than guessing on something with traceability implications and brand risk.

Yes. If your wine cellar is published, the bot reads it and pairs from what's pourable. Region, producer, vintage, and price all read from the list. Pairings match the cooking style (raw, grilled, meuniere, baked) and the price band the guest is comfortable with. It will not invent a 1962 Montrachet that's not in your cellar, so the recommendation is always a bottle the sommelier can actually open and decant.

The bot reads shellfish tags on every dish, and for guests who flag a shellfish allergy it filters the menu accordingly. It defaults to a confirm-with-the-team note even on shellfish-free dishes, because cross-contact in a kitchen that runs oysters, shrimp, and crab through the day is a real risk. For severe allergies, the bot recommends a brief conversation with the chef before ordering rather than relying on the chat alone.

Yes. SleekAI supports multibot, so a flagship raw bar and a casual sister seafood spot each have their own chatbot scoped to that restaurant's catch, cellar, and policies. Cross-location confusion, like a flagship-only Belon No. 2 quoted at the casual sister that does not source them, never reaches a guest. Shared brand-wide notes like the sustainability policy can sit in a shared knowledge layer.

SleekAI is a one-time WordPress plugin license. You bring your own API key (OpenAI, Anthropic, Google, or OpenRouter), so the running cost is just the model tokens, typically a few cents per conversation. For a seafood restaurant doing 200 chat conversations a day on GPT-4o-mini, that runs well under what a per-location SaaS chatbot subscription would charge each month.

SleekAI does not book tables itself. It links the guest to your existing reservation system at the right point in the conversation, with party size, date, and any seating preference (raw bar counter, corner booth, terrace) already captured. That keeps booking data in one place and lets the chatbot focus on catch, oyster, and pairing questions rather than running a parallel reservation channel.