What you don't see is what costs you the most.
The requests that never reach your team. The leads going to competitors before you've even read their message. The sales you think you've lost — but that were never actually counted.
What if we looked at how much it's really costing you?
Already deployed in hospitality. Adaptable to your sector.
30% of leads are never contacted.
According to Forbes. And per an InsideSales study, 50% of sales go to the first company that delivers a meaningful response.
30%
Leads never contacted (Forbes)
50%
Sales won by whoever responds first (InsideSales)
<5 min
Critical response window for conversion
Every business with a high volume of inbound conversations runs into the same problem: a patient hesitating at 11pm about a treatment, a qualified lead landing in the middle of the night, a customer trying to book for tomorrow. What isn't handled in time is lost, and no one even notices.
Three concrete ways to work together.
End-to-end conversational AI agent deployment
I build and operate full automation stacks for companies that want to delegate their inbound interactions. WhatsApp Business, voice (smart IVR), reservation/CRM integrations, payment, human escalation. Setup + monthly retainer.
- For: service businesses with high interaction volume
- Engagement adapted to scope (starting at 6 months)
AI Flash Audit
1-hour scoping call + PDF deliverable within 48 hours. For executives who want a concrete read on the AI potential inside their business before committing to a deployment. I quantify automatable volume, propose an architecture, and give a pricing range.
- For: operational leaders who want a realistic roadmap
- Deliverable: PDF within 48 hours
Custom technical integrations
For companies that have already identified a workflow to automate and are looking for a technical partner. Stack: n8n, Claude/OpenAI/Mistral, RAG (Retrieval-Augmented Generation), vector databases (pgvector, Pinecone), webhooks and custom APIs.
- For: complex internal workflows, CRM integrations, back-office automation
- Stack: n8n, LLMs, RAG, vector databases
Between a demo that works and a system that holds in production, there's 90% of the real work.
That's the part I do.
Case study: Catch Beach Club, Twinpalms group — Phuket
First production deployment for the Twinpalms group.
A premium beach club receiving ~150 WhatsApp messages a day in high season. A single person handling everything, overwhelmed. The result: lost customers, revenue going elsewhere — without anyone noticing.
150+ messages/day
handled automatically in high season
+7 languages
served in real time (FR/EN/ES/RU/TH/ZH/AR)
30 seconds
average response time
Zero
request lost outside service hours
The AI agent responds, checks live availability on SevenRooms, proposes the optimal area, takes payment via Stripe, and escalates to a human manager whenever a case falls outside the playbook. No more requests slipping through the cracks.
Why me?
Before building AI agents, I spent 3 years in premium hospitality: Hôtel Costes in Paris, Cherry and Playamigos (Indie group) in Paris and Saint-Tropez.
I saw from the inside the volume of WhatsApp and phone requests that go unanswered in high season, hundreds of bookings going elsewhere because no human could reply at 11:47pm on a Tuesday.
That ground-floor observation is what gave birth to my studio. I don't sell AI in theory. I solve an operational problem I lived through.
What changes from one case to the next isn't the tech. It's the business context. And that's where the quality of the deployment plays out.
Let's talk?
30 minutes to understand your context and see whether I can bring real value. No sales pitch.