AI chatbot that never sleeps.

What they came in with.

A food brand was losing orders after hours. Questions about ingredients, allergens, catering menus, and delivery areas were going unanswered until the next morning, by which point the customer had already moved on.

The brief: handle the common questions automatically, qualify catering enquiries, and escalate anything complex to the team. Free the human touch for conversations that actually need it.

How we think about it.

Every project starts with the business problem, not the solution. Before touching any tool or writing a line of code, we spend time with the people who own the outcome. We map what's broken, what's actually worth fixing, and what success looks like in practice, not in a brief. Only then do we build.

We audited six months of WhatsApp and email threads to map the 12 questions that covered 80% of all enquiries. Those became the knowledge base.

The chatbot is trained on the actual menu, ingredient lists, catering packages, and delivery zones. It doesn't guess, it retrieves. And when it can't answer, it says so and collects a contact.

The work, in plain language.

• 01Knowledge base built from real customer conversations, not marketing copy
• 02Handoff logic triggers on intent signals: price negotiation, custom orders, complaints
• 03Catering enquiry flow structured to capture date, headcount, and budget before escalating
• 04Response tone trained to match the brand voice, not sound like a bot
• 05WhatsApp-native integration so customers use the channel they already prefer