How we use your text

With "Import itinerary", you paste an itinerary you already have (an AI's answer, a message, your notes) and Ciranda builds the trip from it. This page explains what happens to that text.

Who reads the text

The text goes from our server in São Paulo to Amazon Bedrock, the AI service of Amazon Web Services (AWS). There, an AI model, Anthropic's Claude, reads the text and returns a draft of the trip: cities, dates, legs, lodging and place ideas.

You review the draft before the trip is created. Nothing is created without your review.

The text is only used to build your trip. The model doesn't answer questions or do other tasks: text that isn't an itinerary is turned down.

Where the text is processed

Your data stays in Brazil, in São Paulo. The "Import itinerary" text is the exception: it may be processed for a few seconds on AWS servers outside Brazil, in any country where AWS offers the service. It isn't stored there, isn't used to train models and isn't shared with Anthropic.

If Claude fails, the text may be read by another model, Qwen3 32B, also on Amazon Bedrock, processed in São Paulo.

This explanation is still under legal review. If the review doesn't approve processing outside Brazil, import will switch to a model processed only in São Paulo, and this page will change.

What's kept

What not to paste

Don't paste passwords, ID documents, card details or health information. To build the trip, cities, dates, legs, lodging and places are enough.

Daily limit

Questions

Write to support@awayproof.com or see Help. The legal details are in the Privacy policy.