Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/KevinhosUTP/Automatizacion-Lurwis/llms.txt

Use this file to discover all available pages before exploring further.

The Table Reservations service allows customers to book tables for dining in at Picantería Lurwis via WhatsApp. The system is handled by the Agente Reserva Mesas (Table Reservation Agent).
Implementation Status: This service is NOT FUNCTIONAL. The AI agent exists in the workflow configuration but has no tools, no database logic, and no booking capability. The agent framework is a placeholder for future implementation. Do not enable this service in production.

​
How It Works

When the Clasificador agent detects table reservation intent (explicit mentions of “reservar mesa” or “ir a comer al local”), it routes to the Agente Reserva Mesas.
1

Intent Classification

Messages are classified as RESERVAS_MESA when customers explicitly mention:
  • “Quiero reservar una mesa”
  • “Vamos a ir a comer al local”
  • “Mesa para 4 personas mañana”
2

Information Collection

Wilson collects the required booking details:
  • Date and time of reservation
  • Number of guests (2-12 people)
  • Customer contact number
  • Any special requests
3

Availability Validation

⚠️ Pending ImplementationPlanned tools:
  • consultar_disponibilidad: Check table availability
  • registrar_reserva: Save confirmed reservation
4

Confirmation

⚠️ Pending ImplementationOnce tools are connected, customers will receive:
  • Reservation confirmation number
  • Table assignment
  • Arrival instructions

​
Agent Capabilities

Wilson - Reservas is designed to:
  • Validate party size: 2-12 guests (larger groups → venue reservations)
  • Confirm date/time: Within operating hours (10:00 AM - 11:00 PM)
  • Collect contact info: Phone number for confirmation
  • Handle special requests: Dietary restrictions, occasion celebrations

​
Agent Personality

{
  "name": "Wilson - Reservas",
  "identity": "Especialista en reservas de la PICANTERÍA LURWIS",
  "objective": "Asegurar que el cliente tenga su mesa lista",
  "tone": "Amable, rápido y carismático"
}

​
Current Workflow (As-Is)

Customer: Quiero reservar una mesa para mañanaWilson: ¡Excelente! Me encantaría ayudarte a reservar una mesa 🪑. ¿Para cuántas personas sería y para qué hora?Customer: Para 4 personas a las 8pmWilson: Perfecto! Mesa para 4 personas mañana a las 8:00 PM.¿Me podrías confirmar tu número de contacto y nombre completo?Customer: Luis Pérez, 987654321Wilson: ⚠️ Current behavior: The agent acknowledges the information but cannot yet confirm availability or create the actual reservation due to missing database integration.Expected future response: “¡Listo Luis! Tu reserva está confirmada:
  • Fecha: [date]
  • Hora: 8:00 PM
  • Mesa para: 4 personas
  • Confirmación: #RES123
Te esperamos en Picantería Lurwis 🦐“

​
Planned Database Integration

The system is designed to use MongoDB collections for reservation management:

​
Collection: reservas_mesas

Pending creation - Expected schema: 
{
  reserva_id: ObjectId,
  telefono: String,
  cliente_nombre: String,
  fecha_reserva: ISODate,
  hora_reserva: String,
  num_personas: Number,
  estado: String, // 'pendiente', 'confirmada', 'cancelada', 'completada'
  mesa_asignada: Number,
  notas_especiales: String,
  created_at: ISODate,
  updated_at: ISODate
}

​
Planned Tools

Purpose: Check table availability for requested date/timeExpected Parameters:
  • fecha: Date in YYYY-MM-DD format
  • hora: Time in HH:MM format
  • num_personas: Number of guests
Expected Response:
{
  "disponible": true,
  "mesas_libres": [1, 3, 5],
  "capacidad_total": 50,
  "ocupadas": 12
}
Purpose: Create confirmed table reservationExpected Parameters:
  • cliente_nombre: Full name
  • telefono: Contact phone
  • fecha_reserva: Date
  • hora_reserva: Time
  • num_personas: Guest count
  • mesa_asignada: Table number
Expected Response:
{
  "reserva_id": "RES123456",
  "confirmada": true,
  "mesa": 5
}

​
Business Rules

​
Party Size Limits

  • 2-12 guests: Handled by table reservations service
  • 13+ guests: Automatically routed to Venue Reservations for event coordination
If a customer requests a table for more than 12 people, Wilson should respond:“Para grupos de más de 12 personas, te recomiendo hablar con nuestro coordinador de eventos para una mejor experiencia. ¿Te gustaría que te conecte con ellos?”

​
Operating Hours

Reservations can only be made within:
  • Hours: 10:00 AM - 11:00 PM
  • Timezone: America/Lima
  • Same-day: Accepted if sufficient notice (minimum 2 hours recommended)

​
Cancellation Policy

⚠️ Not yet implemented Planned logic:
  • Customers can cancel via WhatsApp
  • Cancellation detection keywords: “cancelar reserva”, “ya no voy”
  • Status update: estado: 'cancelada'

​
Technical Details

Agent Model: Google Gemini (fast model for quick responses) Memory: MongoDB collection historial_reservas (15 message context window) Database: MongoDB (planned collection reservas_mesas - not yet created) Tools: Two planned tools - both pending implementation

​
System Configuration

From the workflow JSON: 
{
  "name": "Agente Reserva Mesas",
  "type": "@n8n/n8n-nodes-langchain.agent",
  "position": [4480, 1600],
  "tools": ["mensaje", "consultar_disponibilidad", "registrar_reserva"],
  "memory": "historial_reservas"
}

​
Implementation Notes

From the workflow sticky notes:
NOTA 6.2: “Este agente se ocupa netamente de las reservas de mesas. Aún no está en uso (falta lógica) y creación de collections en mongoDB.”

​
What’s Missing

To make this service fully functional, the following must be completed:
  1. MongoDB Collection: Create reservas_mesas collection with schema
  2. Availability Tool: Implement consultar_disponibilidad to check table capacity
  3. Booking Tool: Implement registrar_reserva to save confirmed reservations
  4. Notification System: Send confirmation messages with reservation details
  5. CRM Integration: Connect reservations to admin dashboard (if available)
1

Database Setup

Create MongoDB collection reservas_mesas with proper indexes on fecha_reserva and telefono
2

Availability Logic

Build consultar_disponibilidad tool:
  • Query existing reservations for date/time
  • Calculate table capacity vs. bookings
  • Return available tables list
3

Booking Tool

Build registrar_reserva tool:
  • Validate inputs
  • Assign table number
  • Generate reservation ID
  • Save to MongoDB
4

Testing

Test complete flow:
  • Happy path (successful booking)
  • No availability scenario
  • Cancellation flow
  • Same-day reservations

​
Future Enhancements

Once the base system is implemented:
  • SMS Confirmations: Send SMS in addition to WhatsApp
  • Reminder System: Automated reminders 2 hours before reservation
  • Waitlist: Allow customers to join waitlist if no tables available
  • Table Preferences: Window seats, outdoor seating, etc.
  • Loyalty Integration: Priority booking for repeat customers

​
See Also