Arnau Sanz
Agentic Strategy & AI Transformation LeadKarminaAI StudioAgents · Services · Training
SUCCESS CASE · Restoration
UDON
This case explains the challenge of UDON, the solution WhatsApp Desk and the decisions that allowed it to be applied with knowledge, human control and reviewable results.
StrategyOwn knowledgeProportional permissionsHuman supervisionMeasurement and improvement
01
A WhatsApp flow assisted by AI to guide queries, retrieve valid information and prevent exceptions from getting caught in an automatic response
In restoration, questions arrive urgently: schedules, locations, reservations, letter, promotions, allergens or incidents. The challenge for UDON was to expedite recurring queries without giving generic answers when the information depended on the restaurant or could affect a person's health.
Karmina AI Studio implemented a WhatsApp Desk connected to official sources and a derivation map. The AI identified the need and prepared the next step; the team or the restaurant assumed any exceptions, claims or sensitive queries.
Main CTA: Design a WhatsApp useful and governed CTA Secondary: Discover the WhatsApp Desk
02
Summary Fact Sheet
- Brand: UDON
- Sector: Restorative
- Challenge: orient recurring queries and refer them to the appropriate point
- Principal Agent: WhatsApp Desk
- Complementary agents: Support Desk · Inbox Flow
- Service: digital care and community care
- AI layer: conversational agent, classification of intentions and recovery of knowledge
- Human role: reserves, incidences, allergens, compensations and local exceptions
03
El challenge
Attention, a correct answer in general may be wrong for that restaurant
Schedules, availability and promotions change according to location and time. A consultation on allergens cannot be resolved with an inference, and a claim needs listening and accountability.
The challenge was to automate navigation and retrieval of stable information, keeping a human door visible.
04
The starting point
They unified tokens of restaurants, schedules, booking channels, current menu, frequent and responsible questions. Each data needed an owner and a refresh frequency.
Intentions, escalation phrases and information that the agent was not to ask for or hold were also defined.
A CAS IS NOT COPY
We can understand what is transferable in your context.
In a free video call we will separate the useful idea from the one that depended on the client, the data, the team or the specific moment.
Video call · free · no obligationBook a free video call ↗A short conversation to understand your starting point and bring clarity.
05
The Karmina AI Studio approach
The circuit combined:
- Inbox Flow to classify the query.
- WhatsApp Desk to guide and retrieve applicable information.
- Support Desk Preparing to move in context.
- A person from the restaurant or central team to confirm the exception.
The conversation explained that automatic assistance was being used and offered clear access to a person.
06
How it worked
- Identification of language, need and restaurant when necessary.
- Consultation of the current file.
- Guided response or referral to the reservation system.
- Immediately stopped in front of allergens, health or claim.
- Go through with a summary, without forcing the conversation to repeat.
- Aggregate analysis of questions to improve information and service.
07
Governance and Limits
- The agent did not confirm a reservation if it was not connected to the official system.
- There was no guarantee of absence of allergens.
- It did not settle claims or compensation.
- It collected only the necessary data and for the purpose indicated.
- Changes in schedule or letter required a responsible source.
08
Results
The system reduced repetitive queries to local teams and improved the handover of situations needed by one person. The aggregated questions also helped detect confusing or outdated information.
Main results: faster initial orientation, less repetition for the user, more correctly derived queries and better visibility of recurring questions.
YOUR START POINT
Tell us what you want to transform.
A summary of the challenge, the current process and the expected result is enough to begin to value a route of its own.
09
Related agents and services
- WhatsApp Desk: Conversation and Orientation.
- Inbox Flow: classification of tickets.
- Support Desk: response delimited and scaled.
- Digital Attention: circuit design, contents and human operation.
10
Closure and CTA
Automate frequently asked question to give more space to real hospitality
The team
The team that connects strategy, implementation and monitoring
Ignasi Llorente
AI Partnerships & Business Director
Martina Díaz
AI Finance & Business Intelligence Director
Daniel Mallén
AI Operations & Workflow Lead
Eva Trigo
Generative Creative Director
Ariadna Miranda
AI Administration & Knowledge Operations Lead
Claudia Gonzalez
AI Client Success Manager
Jonan Basterra
AI Social Strategy Director
Laura Sabio
Agentic Social Media Lead
Agustina Córdoba
Content AI Specialist
Alba Gómez
Social Listening Specialist
Paola Cudeiro
AI Client Success Specialist
Roger López
AI Project & Workflow Specialist
Mar Berzal
AI Content Strategy Lead
Andrea Gallego
Community Intelligence Specialist
Candela de la Fuente
Content AI Specialist
Blanca Gómez
AI Community Operations Specialist
Cintia Armajach
Content AI SpecialistLet’s talk
We are talking about a similar challenge.
Tell us about your situation, even if the challenge is not fully defined. A member of our team will respond to understand the process and discuss a first step.
hello@karmina.ai