Instant Lead Replies: Boutique Hotel AI Concierge
Katya Dominguez
The problem
A boutique hotel received wedding, corporate event and vacation inquiries by email and through its website. Each one waited until someone on staff had time to answer.
What I built
By email, 2 agents in n8n:
A parser agent reads each inquiry in Gmail and pulls out the event details.
A reply agent writes an answer in the hotel's voice and saves it as a labelled Gmail draft. Staff read it and press send.
On the website:
A Voiceflow chat answers guests from the hotel's own files and asks about group size, length of stay and budget to qualify the lead.
Boutique hotel AI Concierge: Retrieval Augmented Generation (RAG)
Tools
n8n and Gmail run the email agents. Voiceflow runs the website chat, with OpenAI and retrieval-augmented generation (RAG) behind the answers.
The result
Labor costs fell 40%.
Guests waited 50% less for an answer.
Staff still review every email reply before it goes out.
My role
I designed and built both agents, set up the chat's knowledge base, tested the replies and handed the system over.
Next step
An inquiry agent with staff review fits the Advanced AI Systems Sprint. Start with the Free AI Systems Fit Call and bring 5 inquiries you answered last month.
Like this project
Posted Oct 22, 2026
Labor costs fell 40% and guests waited 50% less. I built n8n email agents that draft Gmail replies for staff to send and a Voiceflow chat that qualifies guests.
