Case study 03 · Hospitality technology
Conversational ordering and kitchen control
A WhatsApp ordering flow connected to menus, kitchen queues, staff actions, customer updates, and feedback.
Working conversational workflow + admin · Client name withheld
- Menu
- Confirm
- Kitchen
Challenge
A busy hospitality team needs ordering, the kitchen queue and customer updates to work as one system. Retyping orders by hand means missed items, unclear wait times and orders that are hard to trace.
Let customers order naturally in chat while prices, quantities, order states and staff accountability stay exact, with a fallback when the AI is unsure.
System response
Chat messages are read against a structured menu, then confirmed explicitly. Orders move to a staff and kitchen workspace with quick status actions, scheduling, payment state, customer notifications and an audit trail.
Product scope
- QR-to-WhatsApp customer entry
- Natural-language menu and quantity parsing
- Structured confirmation, totals, and queue-based ETA
- Kitchen board with order-state controls
- Menu, customer, team, promotion, and feedback administration
- Returning-customer reorder and post-order feedback
System architecture
The chat layer interprets customer intent against a controlled menu.
A confirmation step turns the conversation into an order record.
Kitchen and staff screens manage each order’s state.
Status changes trigger customer updates.
Operations data feeds menu, customer, and feedback views.
Observable outcome
What the product and its repository show. These are not confidential client figures.
Natural conversation without sacrificing transaction structure
One order state shared by customers, staff, and kitchen operators
A reusable pattern for chat-led commerce beyond hospitality
Repository evidence
- Documented webhook and order API surface
- Working staff command and kitchen workflows
- Prompt hardening and state-management test suites
Built with
- TypeScript
- Node.js
- WhatsApp/Twilio
- PostgreSQL
- Optional LLM parsing