AI agents for sales teams in restaurants

Enquiries for a forty-person party land on WhatsApp, Instagram, a venue-hire platform, and the website form, and get answered whenever the events manager comes up from service, by which time the organiser has asked three other places.

→An agent sweeps every channel on schedule, logs each enquiry with its date and numbers, checks the diary, and drafts a reply with the open dates for the events manager's approval.

Every proposal is rebuilt from the last party's document, with menu prices, drinks package, and room hire re-keyed by hand and a minimum spend that someone has to remember to check.

→An agent drafts the proposal from the current set menus and price sheet, works out the per-head and minimum spend, and queues it for approval before it goes to the client.

A confirmed party's numbers change twice in the week before, the function sheet is stale, and the kitchen preps for forty when twenty-eight are coming, while the deposit that held the date is still unpaid.

→An agent updates the function sheet when the client's numbers change, flags the kitchen, and drafts the deposit chase for the sales manager to approve.

Christmas bookings depend on who remembers last year's office organisers, the concierge who sent six parties, and the company that wanted gift cards, so the pipeline lives in one person's head until December is full or empty.

→An agent builds the organiser and referrer lists from last year's bookings, drafts personal rebook notes for approval, and reports confirmed parties against the same date last year.