Skip to content
flylabs.ai
HomeWorkServicesProductsStack
IT (italiano)/EN (English)
Let's talk
← All work
Hospitality · Sardinia

AI-drafted review replies, in the house tone

ClaudeMake.comApifyGoogle Sheets
1000+
reviews analysed
17%
multilingual reviews

The problem

Review replies written by hand, one by one, across languages, on three different platforms: Booking, Google and Tripadvisor.

The solution

An AI pipeline downloads and checks reviews from Booking, Google and Tripadvisor every day, and drafts replies in the house tone: a human reviews and publishes.

Daily pipeline diagram: scrape of new reviews from Booking, Google and Tripadvisor, dedup, draft generated by Claude in the house tone, worksheet, summary email, operator review(opens in a new tab)
The daily pipeline: from new reviews to a ready-to-publish draftTap to enlarge
Weekly loop diagram: current prompt and operator feedback consolidated by Claude, with a prompt update or a no-change email(opens in a new tab)
The weekly improvement loop, driven by the operator's feedbackTap to enlarge

How it went

The client came from a DIY attempt with an unconstrained autonomous agent, with no control over costs or permissions: a controlled approach was needed. Before automating anything we analysed the full historical record of reviews and the owner's replies: from those, Claude distilled a tone and style guide (register, opening and closing formulas, complaint handling), validated by the client on sample drafts before the pipeline went live.

Every morning the pipeline checks for new reviews across all three platforms where the property is reviewed, Booking, Google and Tripadvisor, discards the ones already seen and drafts a reply for each in the review's own language, with an Italian translation alongside. Sensitive cases, such as very low ratings or concrete complaints, are highlighted and never published blindly.

The operator finds everything in a shared worksheet: they review the draft, publish it via each platform's own management console and can leave feedback on every reply.

Once a week a second process re-reads the operator's feedback and the prompt in use: when needed, it rewrites a consolidated version, archives the previous one and reports the differences. The system gets better every week, and every reply goes through a person before it's published.

The product behind this case: Stellar Reviews →

Got a similar process?

Tell us about it in a couple of lines: in the first call, which is free, we'll tell you whether and how it can be done. If it can't, we'll say so.

Let's talk →Next case · WeGrocery: the open-source platform for group orders →
flylabs.ai

Practical AI, built around your work. Then you run it.

HomeWorkServicesProductsStackPrivacyCookie policyinfo@flylabs.ai
© 2026 flylabs.ai
Cesare Bedin (LinkedIn) · VAT no. 05755090288Federico De Cillia (LinkedIn) · VAT no. IT13990330964