Use case — Hotels & spas
Between bookings, the front desk, and housekeeping, information gets lost.
I designed this type of system for Pluto.AI, an AI platform for hospitality and wellness. The principle: the system prepares, a human approves anything that commits the property.
Bookings
36 bookings
In practice
Solutions ready to implement
Availability verified by a human
Trigger
An availability request comes in (room, treatment, service).
Agent action
The system prepares the reply based on what's actually available.
Validation
A human approves before it's sent — the system never announces it on its own.
Result
A reliable answer, never a made-up availability.
A reliable recap, never hallucinated
Trigger
A booking is confirmed.
Agent action
The recap is composed from what was actually booked.
Validation
Nothing is freely generated by the AI — the recap stays reviewable before it's sent.
Result
A reliable recap, never hallucinated.
An upsell that targets the right offer
Trigger
A booking is recorded.
Agent action
The agent identifies relevant add-on offers (upgrade, treatment, option).
Validation
The suggestions target what was actually booked, not a generic menu.
Result
A relevant upsell, never generic.
Method
How I work
-
1
Understanding your business
I take the time to understand how you actually work: your tools, your clients, your constraints — before talking solutions.
-
2
Identifying friction points
Together we pinpoint exactly where things get stuck: where you lose time, where information gets lost, where a customer waits too long.
-
3
Analyzing the need
I look at what, among these friction points, can genuinely be handled by an AI system — and what definitely shouldn't be.
-
4
Proposing solutions
A concrete architecture, wired into your existing tools, with what stays under your control clearly identified.
-
5
Implementation
The system is built and connected to your real environment, not an isolated demo.
-
6
Training on the tools
Your team learns to use, adjust, and trust the system — without depending on me for every setting.
-
7
Testing
The system runs on real cases before being rolled out fully, to catch what needs fixing before it reaches your clients.
-
8
Optimization
Early results help refine the rules, thresholds, and edge cases.
-
9
Final version
A stable, documented system you can keep evolving, with or without me.
Custom-built
Your situation doesn't fit any of these boxes? Let's talk anyway.
The solutions above are a starting point, not a fixed menu. If your problem doesn't resemble any of the cases described here, that doesn't mean there's nothing to do — it just means we start from your case rather than a template.
Positioning
A vision that goes beyond the technical
A poorly designed AI system can hurt your customer journey, your brand image, or your sales process just as much as it can improve them. Before AI, my background came from marketing, design, and sales — I look at a project as a whole (marketing, UX, sales, product, design), not just the technical piece.
Trust & transparency
Trust & transparency
The system prepares, you decide
Anything that commits your business (a rate, a message to a client, approving a file) goes through human validation before it goes out.
Nothing is a black box
You always know why the agent did what it did, and whatever was prepared or sent stays reviewable.
We plug into your tools, we don't lock you into a new one
No extra platform to learn, no forced migration.
See also
Other industries, the same method
Let's talk about your property
A first conversation, no strings attached, to understand where you're at and whether it's worth going further.
I reply within 24 business hours.