Allergen questions in restaurants and food brands have two failure modes. The cautious one is a server saying 'I'll check' and disappearing, sometimes coming back with the wrong answer. The reckless one is a server saying 'should be fine' without checking. Both are bad for the diner and the brand. The data needed for a correct answer is almost always already on file. It is just stored in a kitchen book or a spreadsheet, not the customer-facing menu page.

SleekAI reads your menu_items custom post type, the allergens taxonomy with terms like dairy, gluten, sesame, soy, mustard, sulphites, the ingredient list field, and an optional cross_contamination note where the kitchen flags shared fryers or surfaces. The bot answers 'does the falafel wrap have sesame' with a clear yes or no, the source ingredient, and the cross-contamination context.

For chains that update menus seasonally or by location, display conditions can scope the bot to a single venue's menu so the answer is venue-specific. Logs capture the dish queried, the allergens asked about, the model, tokens, and the originating page, which doubles as a feedback loop for the kitchen on how often a given allergen comes up.

Allergen questions are a hospitality test in miniature. The diner who asks is trusting the restaurant or food brand with their health, sometimes severely. The way the question is answered tells them how the venue treats people, not just food.

A rushed 'should be fine', a slow back-and-forth to the kitchen, or a printed sheet older than the current menu all signal the same thing: nobody owns this part of the experience. A chatbot does not replace the human relationship at the table, but it changes the picture for everyone who never makes it to the table. The diner Googling whether to book, the parent vetting a children's menu, the office team picking a venue for a colleague with a severe allergy - all of them get an answer that reflects exactly what the kitchen wrote down, in seconds, without anyone needing to call.

The dignity of being able to ask the question without a fuss is itself a win. The kitchen-side benefit is just as real. Logs reveal which allergens come up most often, which dishes are routinely asked about, and which substitutions diners actually want.

Over time the menu can be designed with that signal in mind. A dish that produces a steady volume of nervous queries because of one sesame garnish can be reworked to be safe by default, and the new version sells more because the friction is gone. The chatbot is not a marketing widget.

It is the front line of how a venue takes care of its guests.